Niall’s virtual diary archives – April 2002

Thursday 4th April 2002: 11.14pm. It feels like I've been here for weeks, but in fact it's only been four days. Weird. We returned here last Sunday after having taken the previous week off for Easter holidays as it were. I spent the weekend here programming various things, then the first part of the week in Hull and the latter part of the week in Madrid. It all went pretty well, although there was certainly a lack of time. Especially in Madrid, where me watching all of the saved television episodes back to back actually occupied all of the three days there.

Well, three weeks and five days to go. Looking forward to it already, and I have started a list of miscellaneous things I've been wanting to do for ages but never had the free time. Real random stuff. For example recently I after three days of mucking around finally got BeOS to install and it is interesting - although I can't quite see what all the hoohah is. Maybe programming for it is different? Dunno, but it's certainly not as responsive as I had been led to believe. They could have done with more threads

Speaking of which, I have so many ideas for what substantial project I should undertake for my six months of freedom. Something really passionate and good. I'd like to develop a system where data streams are pervasive - I've spoken about that theory before regarding Tornado. One idea was to make the file system unified with the driver model (a la Unix) but furthermore make every process and thread also part of the file system along with every possible point of i/o. I mean, even inter-component i/o should appear in a file system. They would effectively be break out points which could be connected with one to many relationships much like objective C or Qt. Obviously, each data stream would have hierarchical filetyping eg; Data->Image->JPEG->JFIF for example. Or Data->Document->Text->HTML. Or even Data->Movie->MPEG->4. That way even if the filetype is unknown, you at least know roughly what it contains (and you can open it in a binary view).

All files would be stream outputs and all windows would be stream inputs. So you could legitimately connect a file to a JPEGtoBitmap converter to the Bitmap stream input on a window object. And of course this ability wouldn't exist just to the programmer - the entire user interface would be built on the user making connections between data streams. Another thing I had been strongly thinking of given my horror that Qt has no error handling was implementing system-wide structured error handling where the object interfaces have a language portable and mostly transparent error checking and handling scheme. I think error handling standardisation by the system a very good idea.

Right, enough wittering, and definitely time for bed. Good night everyone, hope you all have nice dreams and be happy till our next reunion!

Saturday 27th April 2002: 2.57pm. The last weekend. In only three days, it'll all be over. It turned quite nasty over the last two weeks especially, coming to a head last weekend when personal insults got thrown at me. I get the feeling they don't want me to leave - but then I suppose they feel the world beginning to crash down around them and that breeds desperation. They got called to Munich a few days ago, I believe they told them another pack of bullshit and so they'll struggle on for a while longer yet. But without me.

When it became clear I wouldn't stay willingly, various threats were made regarding removing my social security and indeed portions of the money due to me if I didn't do another month or even two weeks here. I honestly don't know what's going to happen next week, maybe they pay me the two months back pay they haven't as yet plus my bonus or any variation in between. I am not going to let Canada happen again though - if they are so much as a centimo short, I will sue. I would have left last February rather than return here if I thought I wasn't going to get my money. And also, genuinely, I had hoped things were going to turn out better.

While the engineering problems have been great here, the major failing I feel has been in fact management and their complete inability to see the situation for what it is. More and more of late they have increasingly ignored opinions from the engineers, the actual people like me who make it work. They have imposed one final finishing date after another, only to watch each come and go. With each extra promise to the customer broken, confidence in their word disintegrates. With their constant pressure on the engineers to have it all working in X weeks rather than accept it will be X months, an awful lot of corners remain being cut and a lot of time wasted. If they keep at it like this, I foresee the project heading into early next year and sure as hell, I ain't staying around being paid less than the guys doing the wiring for all that long. No, I am absolutely clear in conscience that I have done the right thing in drawing the line in the sand here and now. What I do feel sorry about is the colleagues I have left behind to carry the bucket, plus I have made some good friends whilst working here. I should doubt if I will ever see any of them ever again.

I remember one of my old managers telling me horror stories about projects which went catastrophically wrong. He said he had experienced being an engineer in one once and also said that staying to the bitter end was almost always the wrong thing to do - not just for yourself, who loses the most - but also to management, because so long as you stay they are not awoken to the complacency of fundamental failures within their own actions. When people start deserting you like rats leaving a sinking ship, it can have precisely the right sort of effect in making you challenge your actions and take radical steps to correct the situation. It's weird you know - how many women I have known have perhaps taken similar steps of deserting me with my supposed well-being in mind? In the end, it all swings and roundabouts.

I have been mucking around with the Tornado idea a little more in my very limited spare time recently - we have been regularly working eleven hour days. I have simulated a client-server system by creating threads which is no mean feat to do portably. I also had to write portable pipes which took a while. OTOH, I do now have messages flying between supposed server and supposed client, even if it's not quite stable yet we are moving along. Next thing is to implement data streams, probably one of the hardest things I will have ever written yet as they will be the fundamental glue holding Tornado together. If I get them right, Tornado stands a chance of actually being written. If not, it'll be another failure to chalk up.

Hmm, an interesting challenge beckons. Assuming my money situation has a lot of doubt replaced with certainty soon, I may still be able to pay my own way through the next six months if I dig into savings. I have the computer, I have the ideas and probably will have the free time. All I need now is the discipline and the will to succeed. Only time now can tell what the future holds - a revolution in computing, or another long wait till my next try. It shall be interesting to find out! Meanwhile, expect more diary entries more often given my many more free hours per day, and well, be happy until sometime end of next week!