30 Mar 2006 (Thursday) 

Home again. It seems when I originally planned on going away, I wanted to come back on the Wednesday. Clearly at some point I changed my mind and came back on Thursday instead. That said, travelling on the Wednesday when I felt worst would have been foolish so maybe it was fortuitous. In any case, my tickets were listing the times for the wrong day - didn't matter though, 'cos the ticket's valid for a month, and the train times during the week are the same. Plus just about every train was running late - I was lucky there were big gaps in my timetable.

Return to top | Comment on the diary

27 Mar 2006 (Monday) 

I'm unable to sleep at the moment - this cold and irritancies floating around my head have conspired to keep me awake. So instead of that I thought I'd try to write something useful. I've yet to do this, so instead of that I thought I'd generally rant to myself. It seldom does much good but it beats ranting at other people. That said, I did rant a lot at people on the talker this evening, quite pointlessly.

I had a strange dream last night which involved the sentence 'I picked up the fire axe from the next floor up because I was going to kill the alien robots in the room who were trying to get information out of me, but when I did, the fire alarm went off so I put it back again, realising that the building would be evacuated and so people would be safe', spoken to a policeman who was outside the building after we'd been evacuated. Actually the sentence may have been better (or worsely) worded in the dream, but that was the general idea. Of course, I knew they were alien robots because their cloaks that disguised them as normal people had failed and they'd been left looking like light pink and grey robots rather than people.

At which point I think it's pertinent to mention that on describing the day I had at university when I thought I had a daughter, Ian suggested that I should see a shrink. Hmm. That was years ago, though.

I've just added yet another message to the postponed mail box. It's all quite annoying really.

Return to top | Comment on the diary

25 Mar 2006 (Saturday) 

What else ? Not much really. That's about the whole day really.

Oh, I remember what else happened today; I've got a lovely cut down my nose where Bethany scratched me. It's not sore or anything. It just looks a bit silly.

Return to top | Comment on the diary

24 Mar 2006 (Friday) 

You've got to think about all the things you like and decide whether they're worth sticking around for. [ Doing stuff; Rube; Dead Like Me ]

It just bothers me a little. If I was to die, what would I miss. I have no idea. I keep thinking that maybe I'll find myself an answer, but I've not yet. Maybe at some point.

Return to top | Comment on the diary

23 Mar 2006 (Thursday) 

One thing I have noticed whilst here in the last 10 minutes is that browsing the 'net sucks. Without the filters to get rid of the cacky adverts, sites are just littered with junk. Maybe more normal people have higher tollerances of this type of rubbish. Actually, I know they have - even Alex and Joseph seem to be happy to clutter their own websites with adverts. Somehow it seems to be acceptable to fill pages with random adverts. Of course I now ask myself why I'm surprised by that and whether I feel that it's naive to still think that sites should be cack-free.

Return to top | Comment on the diary

22 Mar 2006 (Wednesday) 

Well, I'm away for a week from tomorrow. I'll be up with Claire and Justin for Bethany's christening. That'll be nice!

Today has been bitty. It was split into two major chunks. One of which was the getting packed and stuff. The other was trying to modify CMunge for some things I happen to need - maybe this time the tool will be 'complete'. The first part of changing things went quite smoothly. Rip out large chunk of code, make generic, put back everything using generic form and test. Nice and easy. Well, not exactly easy, but not too bad anyhow. The second bit was less easy because although I knew what I wanted, the code wasn't quite in the right shape for it. I'm thinking that I went about a lot of that work the wrong way. Fortunately I can revert to the last working copy and just start again. I think that'll make life a little easier. I've also got to work out the 'right' way to return either an error, or a pointer to a block of text, or a null. I'm not sure I really feel a 'good' solution to that, but I'll think about it some more and maybe something nice will happen.

Return to top | Comment on the diary

21 Mar 2006 (Tuesday) 

Mum found me some Orange Chocolate Leibniz. They're quite yummy. And very similar to Clubs for some reason.

I've just had one of those 'spark of genius' ideas that will actually be incredibly useful. It's not actually all that much of a spark, more of a bubbling really, 'cos it's an extension of a number of ideas I've had in the past but this time I really think I've got it. I need to sit down and sort out some details now, though. Probably get to do that on Thursday, I think.

Return to top | Comment on the diary

19 Mar 2006 (Sunday) 

Today's not been too bad. I've been spending it helping someone out building things, and using the spare time to do unimportant, but longstanding things that I've needed to get done but been putting off. Actually I got more done than I thought I would really which is quite nice. And I've fixed a few more simple bugs that needed to be dealt with.

I was looking forward to Midsomer Murders this evening, but it wasn't on. Bother.

It's frustrating that these bugs I keep finding are still there. They're not all mine, of course, but many of them are really stupid. things that people should have caught long ago.

Return to top | Comment on the diary

17 Mar 2006 (Friday) 

I was going to say something useful but then I watched a little bit of Danger Mouse and my brain's turned to custard. Whilst watching the Custard episode.

Return to top | Comment on the diary

16 Mar 2006 (Thursday) 

Playing Darwinia is surprisingly addictive. Whilst doing so, the phrase 'like herding cats' often floats through my mind. Moving the little darwinians around is not all that hard 'cos they do obey orders, but in the absense of orders they do tend to wander. However, I think the thing that made me think that wasn't actually something I realised until much later. When the officers (the darwinians that you've promoted to give those orders) get hurt they let out a little 'mew' which does actually sound like a cat - so much so that twice I've looked around to see if Grendel was around.

And one for the Conspiracies file... Chris Williams (drobe) has been doing some proper journalism for a bit - the kind that doesn't involve just posting something on a website - and I noticed recently that there's a writer on The Register (oh the shame to say I've been reading it) called Chris Williams. Hmm.

Return to top | Comment on the diary

15 Mar 2006 (Wednesday) 

I had a headache all of today. A real 'this isn't going to leave me all day' one. I hate those. Well, obviously. I mean, it's not like you're going to say "Oh, a headache, I've not had one of those in ages, and I've been so looking forward to it". Well, not unless your particularly masochistic anyhow. I wonder if that ever comes up. I mean you've got your regular masochism, but are there people that listen to the same music that they hate over and over again just so that they get the pleasures of a headache at the end ?

I think that's a scary section of society that I'll try to avoid. Not, of course, that it makes them bad people or anything. It's just the idea of intentionally getting a headache for pleasure seems stranger than even I can handle.

And, changing the subject entirely, I ate a whole bar of chocolate today.

Hmm. Thought I had more to say about the day than that. Bother.

Return to top | Comment on the diary

14 Mar 2006 (Tuesday) 

Many annoyances today. Nothing particularly important, but just your regular annoyances that make you want to rip the heads off people and... I dunno... do something that's worse than having ripped their heads off, I guess.

On the plus side, I've done some pretty neat improvements to make the system slightly less laggy on startup, and I got a gorgeous picture of Bethany from Claire.

Why is it that I always seem to be playing catch up ? The 'important' things never seem to get done.

Return to top | Comment on the diary

13 Mar 2006 (Monday) 

My dream last night, as I remember it, involved sitting at the bottom of a waterfall with Caroline and having absolutely no idea what she wanted. Or what I wanted for that matter. Except that being there with her was nice.

At the start of today, I felt a bit 'urgh'. As the day went on, I felt a bit better. Why ? Because of some feeling that things were going better. Unfortunately that's all it was, and when I realised that things went downhill like something that rolls quickly down a hill. It's not a "um, I don't have a daughter, do I?" moment, but it's close-ish.

As a result, I've got significantly less in the way of nails now. I was going to try to get to bed early, but I was so annoyed with myself for feeling good over nothing that I decided to do some other stuff instead.

Return to top | Comment on the diary

12 Mar 2006 (Sunday) 

The word 'Capricious' popped into my head today. Nice little word.

Over the last few days I've been sending out test versions with the subject line 'Build from today', or variations on that. Now that I've had some replies it's blatantly obvious that I'm stupid. Obviously that gives me no indication of which version people are replying to. Doh!

I thought I was out of disc space, but it turns out that I'd got 900M of redundant sources just lying around. Getting rid of that lot feels a whole lot simpler. And it reminds me how important it is to check what you're copying before you blindly copy source directories around.

Why is it that computing cack is so much easier than... well, almost anything else really ?

Return to top | Comment on the diary

11 Mar 2006 (Saturday) 

Chatting on the talker today - for the first time in ages (well, with the exception of last night) - someone mentioned a fun sequence of numbers, and so I just had to go and write a whole load of code to make them. Interesting and fun - something I've not done for quite a few years. Probably since I was at uni. Ah, simple pleasures. The maths may have been simple, but it was mildly amusing anyhow.

Last episode of Stargate SG-1 for this series tonight. That was... different. It's going to be very interesting to see how we get out of that in the next season. One thing that struck me, though was the Ori weapons. They're slow. Projectile slow. Not like lasers, or the more usual energy weapons that the Goa'uld use. Maybe that's just 'cos it looks prettier on screen. Oh, and the Earth ships (again) seemed to be hiddeously outgunned. It seems insane for them to be using their regular projectile weapons against other inter-stellar ships with shields and hugely powerful weapons. I think it's mostly the idea that the best thing that they can use are these same little missiles that they use on earth - against ships that are centuries ahead of them in technology. Seems... mismatched. Of course, sticking said weapons on to an inter-stellar space ship is similarly mismatched. Ah well.

Not a lot else to say today. Much time being annoyed and generally cursing myself. And a little bit of revelation to myself. I was going to rant a little, but for some reason the laptop took about 15 minutes to start up and by the time it had I'd completely lost the will to live. Well, not quite but couldn't remember what I wanted to say. And still can't.

I remember something about the things I was thinking about earlier. There was something about 'a prejudice is an assumption you don't realise you're making' and trying to apply that to things that you don't usually think of like that. Like an assumption that you're not going to spontaneously combust being a prejudice. I think that's why that definition is wrong - 'cos clearly I'm not prejudiced against spontaneous combustion. Um. Or maybe I am and that's the point - it's just not important. Unless you happen to be pro-spontaneous combustion, and the belief that you wouldn't even be pro-that is the prejudice.

However, I think I'm just rambling and stuff. Must sleep.

Return to top | Comment on the diary

10 Mar 2006 (Friday) 

I was trying to write a little caching routine for some data I'm working with, and decided to look at the output, just to confirm that it's actually sensible. I usually do that when I'm trying to make the code efficient just to ensure that there's nothing glaringly stupid about the code. Things like accidentally making a very inefficient loop are easy to do if you're not paying attention and blindingly obvious when you look at the assembler. I was quite amused - and impressed - that the compiler had made such an efficient routine. So efficient that it needed no APCS stack frame - it was completely implemented within 5 registers (R0-R3 and R12).

There was really only one small change that I could make to the C code - a simple change to an upper case check. In that case, c>='A' && c<='Z' is equivilent to ((unsigned long)c-'A') < 26. Of course, the latter is somewhat nasty. The gain is not huge, and it's probably only related to this version of the compiler that can't spot that a range check is being performed. The gain is that you can avoid a pipeline stall because the compiler builds two branch instructions around the code, whereas with the single comparison version only one condition code is necessary and so the code can run inline. Yeah, it's not a huge gain, but since this is part of the inner loop of the code for the caching, I'd at least like to try to get the code 'nice'. Optimising for a compiler is bad, but once in a while it's fun and gives you something useful. The code's not much less efficient than if I'd written it by hand in assembler in any case.

Actually, one of the worst parts of the code is that it has to multiply by 212 which ends up as 4 instructions. If I waste 704 bytes I can use a multiple of 256 and save 3 instructions. Not sure that's worth it to be honest though.

One mildly interesting thing I did notice was that the compiler didn't spot the option to use an 'increment before' instead of ADD r0,r0,#4 : LDMIA.... It's not a particularly common thing to do, but it was still a little surprising to me 'cos I'm sure I've seen it do that before.

I've finished watching Enterprise now. It's been my 'I can't sleep, so I'll watch that' thing for the past few weeks. How can I summarise what I think of it ? I think "I don't really care" sort of covers it. Some bits are good, some bits fun, but the thing that appealed a bit initially became increasingly annoying. No, that's not quite right. Initially, I liked the fact that from episode to episode we had a rolling development - what happened in one episode spilled over to the next. That's fine and I continue to like that. Mainly that's because the general reset switch that got pulled at the end of episodes annoyed the hell out of me in The Next Generation (although less so in The Original Series). Let's not talk about Voyager for its complete lack of continuity in that regard. DS 9 was good for development, but it had to have development because they didn't go anywhere - that's partly why I didn't care so much for the ongoing story line. But in Enterprise it was just annoying. I can understand that the point was that they were on a mission (whatever the mission for the segment was) but aside from the first series and part of the second, the exploration was almost zero. The examination of cultures, morality, and generally 'finding stuff out' was lost and it just became... dull.

Which, I have to then compare to Stargate. There is, and has always been, an ongoing mission for them, although it seems to have become more a prominent part of each series. The replicators season, although we knew for two seasons (I think) that it was coming, became so frustrating because of that. It's that feeling of "Oh, they're fighting blah; how are they going to get out of that ?" which just became a monotony until "oh, I don't care" sets in. The exploration, the discovery, and... the science fiction, basically, seem to have got lost sometimes (significantly in Enterprise Season 3 and 4, but to differing extents in SG-1 seasons 7, 8 and 9 - earlier ones didn't seem to be as prominent). Nothing wrong with having a decent recurring story, but it seems to have been dealt with better such that it's not to the exclusion of the stuff that I like.

Anyhow... that's almost all I'll say about it. Almost. Just a bit more . Enterprise seemed to get very messy at the end, and felt badly thought out for the last few episodes - as if they either didn't have enough content to fill and so padded it out. Reminded me of the end of Angel which felt messy as well.

A few sections are of note, though. The explanation of Klingons more human appearance in The Original Series was an interesting and actually quite neat thing, I thought. Especially as it had been commented on in the DS9 episode 'Trials and Tribble-ations'. That seems to have fit rather nicely with things.

'These are the voyages', the last episode, may be the worst episode of any Trek ever. I'm not sure what else there is to say about it, but it really was incredibly poor.

'In a mirror, darkly' seemed quite utterly pointless to me. All the way through I was expecting some form of cross over to the real universe which we could actually relate to. There was none, and so it was completely and utterly a throw-away episode. However, that's partly because I didn't see so much of DS9 and had completely forgotton 'Mirror, Mirror' from The Original Series. DS9 seems to have had a lot more contact with the Mirror universe, and 'Mirror, Mirror' - which I did actually remember seeing after I had seem a little of it - had set that in motion. The lack, though, of any actual contact with the real universe (the Defiant came across to the Mirror universe, not vice-versa) meant that it didn't actually have any particular meaning at all, and it wasn't the characters or the universe I cared about.

All in all, it was barely 'ok'. In general pretty immemorable.

Really, that's it. Honest.

I've burnt my mouth on a hot chocolate. Which is stupid, 'cos the clue's in the name. It's Hot. Duh.

Return to top | Comment on the diary

9 Mar 2006 (Thursday) 

Once in a while I get emails for Justin Fletcher. Not for me, but for the presenter of a few CBeebies shows Tikkabilla and Something Special as well as doing some touring as well. It's really frustrating to say "No, I'm not him, I'm afraid", especially when they might be children (who, presumably would be sending emails under parental supervision ). So, I thought I'd actually try to contact him myself. I have no idea how easy that will be, but I thought I'd try.

I've dropped a few emails to places that might be able to point me to something. Even if I can just give people an answer of 'try the CBeebies forum, they're friendly there' that would be nicer than just a 'sorry, not me' type response. We'll see what the response is anyhow.

Return to top | Comment on the diary

8 Mar 2006 (Wednesday) 

Lying in my bed, I hear the clock tick and think of you Caught up in circles, confusion is nothing new [ Time After Time, from Twelve Deadly Cyns And Then Some, by Cyndi Lauper ]

I ache a lot today, but I am going to try to sleep more today. Honest.

I was going to go to bed tonight and write a huge long rant. But I really can't be bothered. I'll just summarise the whole thing with "Am I really just meant to understand ?"

Return to top | Comment on the diary

7 Mar 2006 (Tuesday) 

Lots of testing and not a lot else today. I wrote some very tired things last night, one of which might make its way to the 'Stories' section. It's really just a paragraph, but it seemed suitable. Odd though.

Return to top | Comment on the diary

6 Mar 2006 (Monday) 

Over the past week or so I've been having one of those 'afraid to sleep' problems. I'm trying to get over it, but each time I do, it ends up being late morning and then when I actually get up I feel bad 'cos I've not slept enough. But then at night I'm not tired. Bah. Oh well.

Not a lot else to say today. I was going to say it felt like a 'marking out time' day, but I've actually got a reasonable amount done despite it feeling like an unproductive day. Really not a lot else to say.

Return to top | Comment on the diary

5 Mar 2006 (Sunday) 

A while back - dunno when, but in the last 6 months or so - I came across one of the 'distributed search engines' requesting something of my site. I don't remember much about it but I do remember that they exist. And clearly they're in use. What surprised me a little was how this appeared in the logs. It's not often that I get requests from US Department of Defense sites, so it jumped out like a sore thumb when there were nearly 20 requests one after another going through the site (from 204.222.143.129). About 15 seconds later, the requests continue, but this time from a site in the netherlands (81.171.22.189), and then a few seconds later they continue from another site (192.138.77.36), this time a US army machine. And finally they return back to the netherlands machine.

It's pretty neat that a distributed search of the site can take place in the first place. That's pretty funky and we like the idea. But should we be concerned that this sort of thing is running on machines that are inside military networks ? For all I know, the machines may just be the connectivity within the barracks or something, but it seems like an interesting thing to be running. Let's assume for a second that it is inside general military housing. Let's further assume that there is an agreement in place for those users that they won't do anything naughty, like bringing the military into bad repute, visiting websites used by known terrorists or looking for the best ways to import cocaine into the country. You know, those regular things that you try not-so-hard-cos-you-don't-do-that to do.

And let's say that John Q. Badguy, sitting on his machine in Badland starts doing a search, looking for his bondage porn, and searching for guns to help him get his drugs into the country. Nice little distributed network farms this search out to various people and starts off more searches based on the results (keeping up with current searches, refreshing caches and all that). A few of these get passed on to our nice little machine sitting inside the military base and so all said those dodgy sites it hits are fetched, indexed and otherwise brought into the general distributed database. John Q. Badguy's got what he wants, and as a bonus he's just implicated this other little person who just thought they were helping. Hey, that's no big deal because getting a connection between John and this little old machine is next to impossible. But maybe there's a log at the military base - its proxies, or general monitoring kit - and maybe this sort of activity rings alarm bells with someone. Or maybe some group raids the server site of the gun selling people, the bondage porn site, or the "How To Traffic Cocaine In 5 Easy Steps FAQ" and take all the server logs. Or they just serve the ISP with whatever warrants are required and all those logs are just handed over.

Poor little owner of this machine is now being investigated for all sorts of things. That's ok, 'cos they've got nothing to hide. Except that these computer records from the site show that their computer did actually visit those sites. Maybe poor little owner claims that they didn't visit those sites and that the evidence must have been doctored. What about the logs inside the 'secure' military base that kept the same details ? Maybe they could argue that some program had infected their machine and done these things without their knowledge. But what if they installed this distributed client themselves ?

I'm not sure where I'm going with this, but I suppose it's just one of those things you have to be aware of when you install software. "What could it do ?" Obviously you think that about dodgy software, but how much danger is there in something who's entire purpose is to make the general search better ?

Many strange, and possibly unanswerable, questions. I doubt many people worry about this sort of thing.

And I can still move, but I don't speak about it. Pretend I'm crazy, pretend I'm dead. He's too scared to hit me now, He'll bring flowers instead. [ Island, from Oyster, by Heather Nova ]

Annoying stuff today. Testing things against linux worked interestingly well, but testing against RISC OS failed. Which is odd. Code looks right. Doesn't work. I need to go through with some diagnostics to see how far through the system messages are getting before it throws up its hands in disgust.

I was just thinking... "One day I'll be able to not think 'One day...'" which was nicely recursive...

Return to top | Comment on the diary

4 Mar 2006 (Saturday) 

There's nothing left All gone and run away Maybe you'll tarry for a while It's just a test, a game for us to play Win or lose, it's hard to smile [ Make Me Smile (Come Up And See Me), by Steve Harley And Cockney Rebel ]

'Come Up And See Me' is currently being used on an advert for... um... something or other, but I'm not sure what. In any case, it's trying to give a summery, bouncy feel to something or other (is it for a car?) with the track. Only, whilst the music is pretty bouncey, I don't think the track is. I'm not sure, but I've got this feeling it's more "stuff went wrong and you left... but it'd still be nice to see you". At least, that's what I get from the lyrics. Not bitter; just a little sad. Before I looked at the lyrics, though, it seemed more like an 'end of relationship' type track - particularly with the "there ain't no more, you've taken everything", but I'm not really sure it's about a relationship. I think maybe it's more about a group of friends losing touch. Except there's the ending 'don't say maybe you'll try to come up and see me, make me smile', which seems a little bit more bitter, I guess. In any case, whilst I'm not sure of the actual intent, I'm certain that it's not as bouncey as it seems on the musical surface.

And I don't want the world to see me 'Cause I don't think that they'd understand When everything's made to be broken I just want you to know who I am [ Iris, from Dizzy Up The Girl, by The Goo Goo Dolls ]

It's so easy to be caught up in the same music all the time, so it's nice when someone (Andrew in this case) reminds you of music you like but had forgotten about.

Ooh, I've just found that 'Stars' by Lacuna Coil is actually a cover of a track by Dubstar. It's interesting. They're not too much different in general, and I think I can say I like both versions. I'd probably say that the Lacuna Coil version feels a bit more bleak than the original. Actually, that's probably not the right word. Possibly because of the fact that I heard it first, and because I know it better, I could easily imagine the Lacuna Coil version being played whilst sitting and watching the stars on a mild autumn night. Yeah, I think that's probably just by directed association.

Return to top | Comment on the diary

3 Mar 2006 (Friday) 

A few days ago we were watching the end of Due South - an episode I actually vaguely remember seeing - and one of the scientist women working there was Amanda Tapping (Sam Carter, from Stargate SG-1), which was a little surprising. Tonight, I've been watching Star Trek The Original Series 'The Space Seed', which I remember seeing before (probably when I was at school) but was amused by because it leads into Wrath Of Khan. It had no names I knew anyhow, but following on from that, I watched City On The Edge Of Forever, which I thought I knew (but didn't properly remember). The woman's name, Edith Keeler seemed very familiar to me as being important, but it seems it's only important in this episode, which amused me quite a bit. More so that she's played by Joan Collins which at least explained why her face was familiar. But at the end, I was even more amused that the story was written by Harlan Ellison, whose name I do know as an author.

It's not often that I actually recognise names and faces, so that's been quite nice.

Return to top | Comment on the diary

2 Mar 2006 (Thursday) 

Reading through a random diary earlier today I was struck by how little I whitter about computer stuff. At least, that's how I see it. Claire'd probably disagree but it's not really something I'd want to keep a general record of. CVS does that. These days I don't tend to go off on rants about things that people said to me, or said in general on talkers, news groups, or in real life. I'm not sure that's any better because I still get wound up sometimes, but at least it's a better policy in general.

It's sort of similar to the thing I decided about usenet. People that post on usenet a lot either have a lot to say (and usually have no clue about what they do say) or just like the sound of their own voice. And that's fine. But it shouldn't bother me. And every little thing that gets posted doesn't need a response because somebody's bound to chime in with their own, ill-informed opinions (yeah, they may be entitled to them, but that doesn't mean that they're any less wrong if they disagree with me ). So I decided not to post replies to anyone. Unless it was worthy of one. And to be worthy, that has to mean a considered, correct and valuable response. The idea is that if I say something, it should be to respond correctly and fully, and not to correct grammar or the latest misunderstanding of flags on an obscure API that has been around for years people have only just bothered to pay any attention to. Hmm. Yes, ok, there's a little bit of a rant in that sentance. Oh well.

In anycase, this usenet policy seems to have worked well. I don't get wound up with anyone disagreeing with me - because I don't say anything normally. And if I do it's long, correct, and generally 'worthy' . You've got to admire the simplicity. It's almost as good as "la-la-la" can't hear you, except of course I can hear, but just keep quiet. Plus, of course, I only read the technical groups, 'cos people's illfounded opinions and general ramblings about their latest car don't bother me, and it's a waste of my life to read about.

Of course, people that post questions that I can answer get replies by email. Usually they're quite technical questions, or questions about things that I've done or know a lot about, but which it isn't worth posting a public reply ('cos it's only going to be a few lines). Which is why, if you post to usenet, it's vitally important to use a valid address. Otherwise you're pretty much guarenteeing you won't get a reply. From me anyhow. Of course, I expect many people think that's a good thing.

Considering the content, alt.music.lyrics is quite focussed. It's impressive really. Which is an aside thought, but possibly true.

Anyhow, whilst stumbling through these odd little links I began to wonder how many projects that I embark on actually make it to a real use. I'm pretty sure I can't quantify the number, 'cos there's just so much cack that's been put on one side - very little is ever abandoned 'cos that would be too much like throwing something away or (horror) admitting I Was Wrong (yeah, I know I'm mostly ok with that with other people, but being wrong to myself is just awful... hmm... maybe that's why I feel bad? Oh well).

Many projects are put on one side because they expand explosively. Rather than reaching their initial goals, they reach a critical point where they have become useful and the goal is in site and then the goal posts move a few lightyears. The way it usually happens is something like this...

• Decide where we want to go - the thing we need at the end, which is pretty useful and probably required for a few things (most component developments are useful building blocks rather than end-applications).
• Develop around 40% of goal and see useful avenues which might be explored.
• Explore several small developments, discarding about 90%, and realise that the remaining 10% is very useful but part of the design does not fit.
• Discard part of the original goal, to allow it to encompass new avenue. Discard part of the present implementation as it doesn't work generically enough for the current goal.
• Decide that parts of the component are reusable elsewhere. Split them off into a separate library component.
• Make library component more generic, 'cos otherwise what was the point in splitting it off.
• Develop much further. We're now about 80% to our original goal; sadly this is only about 50% to the current goal.
• Decide that goal is too limiting because of environmental constraints. If we change the environment then we can expand the goal's usefulness by an order of magnitude. However, this means changing some of the assumptions we made in designing our goal (eg "must interwork with C" expanded to "must interwork with everything")
• Put on one side because changing environment means that the whole needs to be re-thought.

Of course there's still a huge body of work done by this point which is nearly useful but not quite. Taking the Djinn project (I think that was started around '99 time?) the splitting off of libraries was done and we have some useful bits that came off that (the ServiceList module's second incarnation is a direct offshoot from that), but the expansion of the environment was caused by my recognising that I was re-writing the same code again and again with next to no changes and very little real benefit. The reason the environment changed was that this similar code could be refactored but would require much more information to be generated to do so - from around 20 or so=A0other components offering a introspection interface that could generate it. Which became quite daunting and until it was done it was pointless to pursue the project.

Other times, the project is 'complete' but abandoned because it failed when complete - it should have worked fine and I was happy but it failed in such a catastrophic way that I was too scared to look at it again. The most obvious case of this was the ADFSCache module which I was happy to put to general use after significant amounts of testing and then it trashed my harddisc (ok, just the disc map, but still it hurt). Since then, the source has sat happily on my harddisc (which was recovered, 'cos I wasn't that stupid) but hasn't been touched hardly. At one point I updated it to support sector operations (it was written for the A5000) but aside from building it and examining the code by hand, it's never been run. It still scares me. Possibly because of the fright it gave me.

I guess that's just one of those things that reminds you how vital it is to not be complacent about code. Which brings me back, nicely to the fact that there's not all that much that makes it to completion when I'm happy with it being safe.

Thinking back over the past 4 months or so, most of the projects I've embarked on have come to a solid, useful footing. Off the top of my head there are about 9 projects (for my own note I'm thinking of the codenames PBTS, Stripped source, RTC, DiagData, APCSBT, Interactor, BTSDump, ExtAIF, KInput) which have been successful and (approximately) complete. However, a few of those are offshoots from other, pending projects. APCSBT, and indeed large components of BTSDump and PBTS are offshoots of the PT project, which also gave rise to the DecodeDTF project which was interesting but hasn't found a useful place yet. Strictly 'Stripped Source' hasn't found its place either, but it's complete in any case. Which means that, over the last 4 months or so, we've had 11 projects that I can categorise (I'm ignoring those that are still discussion documents, vague ideas, or a smattering of unconnected sources that haven't built yet), of which 2 have stalled. To be conservative, I'll say that 'Stripped Source' is stalled as well, and make it 3 out of 11. So approximately a third of projects stall - but it's important to remember that PT (stalled) gave rise to APCSBT and large sections of PBTS and BTSDump, which itself spawned DiagData and Interactor. A large chunk of PT moved out to become a library that comprises a lot of APCSBT, and was used within BTSDump, and BTSDump itself was split up into a separate library that's used by BTSDump, PT, and DiagData, and another library from BTSDump became Interactor. Actually the library that became Interactor was completely by design because I wanted to give some example code for DRLink's Linker-sets. So maybe it shouldn't be counted at all. If we take that as an example though, we have 3 out of 11 projects stalling, but 1 of those 3 spawned 3 other, completely successful projects. If we take that project away, then, we would have 2 out of 7 projects being stalled. A little over a quarter of projects stalling. Whilst that sounds better on paper, I feel it's more realistic to consider about a third stalling.

Does that mean that I'm not designing things right ? Or that I'm not carrying things through to their conclusion ? Or maybe it means that I'm being more experimental with my projects than I ought to be ? Should I focus more on the dead certainties than trying the obscure in the hope that they'll be useful ? That's not fair, PT is useful, it's just not finished to a .... um... useful state. Hmm. It will be useful.

I don't know. Since many of the times, the specification is entirely in my head, the general goals are so flexible that knowing that something is complete isn't necessarily easy. I think that makes for some poor engineering at times. Maybe it's a little too dynamic. Of course, the amount of bug-fixing and tracing of general faults takes up around the same amount of time as these sorts of projects in any case.

Like today; I had the best intentions of investigating a particularly long running, difficult to reproduce, but potentially important, fault. Unfortunately, in doing so I uncovered a whole load of little problems that branched off and highlighted other - more serious - problems. So a whole day devoted to just investigating an almost completely unrelated which I can't do anything about. All the diagnotics have gone to the relevant group for their investigation, but it's somewhat disheartening to have spent the whole day doing anything but the thing you set out to do.

In other news (!), I've got nails. Ish. They're not much but I've realised that scratching yourself when you're not paying attention is more likely to lead to cuts if you have nails. I've got a silly cut on the side of my nose, just under my glasses where I wasn't paying attention. Actually, think it's cleared up now, but it was annoying. And typing on this laptop is strange. I notice how I hit keys with the tops of my finger, not the... um... bit with the fingerprint... what's it called ? pad ? or is that only for the palm part of animal's feet ? Anyhow, they slide a little.

And with that revelation, I'm going to go sleepy byes...

... oh, hang on not quite... I found something today - lots of diagnostics means repeated reboots to make sure data is clean - How is that 'So gerph' ?!

The world's a funny place. Guess it's good I don't live in it.

Return to top | Comment on the diary

1 Mar 2006 (Wednesday) 

Woke up late today and missed an appointment. Actually that's not so much due to waking up late as to forgetting what time the appointment was at. Damn. Oh well.

I was talking to mum yesterday, about how strange it was to hear an accordion on a track - Peter Sarstedt's 'Where do you go to my lovely' - and that you wouldn't find that on much music these days. But this morning I remembered that of course Counting Crow's 'A Long December' has an accordion on. Not quite so prominently, but still.

So look into my face Marie-Claire And remember just who you are Then go and forget me forever But I know you still bear the scar, deep inside, yes you do [ Where Do You Go To (My Lovely), from The Peter Sartedt Collection, by Peter Sarstedt ]

Always seems to me to have a tinge of bitterness in the track. I'm not really sure, though. It's just a tinge. I like that word.

Whilst I remember, as well, I worked out the path I took to reach MobuzzTV a little while back... Julian mentioned, a while back, a T-shirt on cafepress of a dinosaur (part of the Flying Spaghetti Monsterism thing), which took me to the Church of the Flying Spaghetti Monster again, and from there the links at the bottom led to the MobuzzTV site. Yay. There were also paths followed to lego churchs at the time, too.

Was that really all today ?

I've been sending out large archives today - many times, because certain broken mail filters are 'detecting' a virus in it. It must be broken because there's no way that my stuff contains a virus. I certainly don't know of anything that would infect a 6M RISC OS archive which was built earlier today from sources. It's not like any of the stuff I'm sending has any extension which other systems would identify as being a potential hazard. Anyhow, it was rather irritating.

Fed up today, firstly for missing appointment and secondly 'cos of just being fed up at things. I said I wouldn't be annoyed but I was. Grr.

I wanna shout about it, But I keep quiet about it. [ Someone's Daughter, from Trailer Park, by Beth Orton ]

Please say I can leave Try to ease this pain that I feel may be killing me Free me from a life That is cold, that is mine [ Somebody's Daughter, from Great Expectations, by Tasmin Archer ]

Randomly connected by having a similar title, firstly because I wanted to find something that expressed "I want to shout about it", and latter just seemed similarly appropriate. I blame mum for playing music yesterday. Yeah, that'll be it. Right.

I should collect a list of pages that are genuinely interesting and informative, to put together for showing people. However I can only think of two at the moment - one is the very cool orders of magnitude book and the other is the one I've been looking at today. It's a periodic table of elements table. An actual table. That's quite cool and would be worthy of note, but there's also loads of other information about the samples the guy has of the different elements and pictures and stuff. It's just quite... cool really. Makes you want to go out and get your own element collection.

And if that sounds to you like something that would be said on MobuzzTV then I might agree with you that it's catching. Don't worry, I'll get over it and I'll be clear of other people's nuances in a few days. Actually I'm writing this like I'm publishing this rather than for me. That's bad. Bad Justin.

In any case, the table was found from the Periodic table of Desserts. Which is cool, in a fun sort of way. Which came from MobuzzTV. I'm so sad.

Return to top | Comment on the diary