Just when I stopped looking, PuTTY has been upgraded. If you are unlucky to be on a Windows box, this tiny little tool could well save your bacon. You get: a Telnet client; a SSH1/2 client; SFTP; a Key generator; a key manager tool; and a command line tool for use with other programs that need to use a SSH connection.
This version now comes with a Unix version, and a Mac OSX version is also in the pipeline. Given that any sane Unix/Linux platform should have OpenSSH on it already, this seems a little pointless, but as they say in their FAQ, some people like PuTTY and did want a Unix version, and the process of porting it to Unix has helped them improve the code, so the Windows version benefits as a side effect. I don't have gtk on my box so I couldn't compile the GUI bits, but the command line pieces worked out of the box on Debian stable.
Like most people using a sane browser I've become very use to using browser tabs. All the good browsers support the idea, Mozilla, Opera and Safari, obviously IE doesn't, but then it's stuck so far in the past, I surprised it's not steam powered.
Today I installed the Mozilla Tabbrowser Extensions, I've know about it for some time, but not bothered adding before. I yanked the xpi file off the 'net and installed it in my Firefox Browser. Seems a very useful addition so far, I'll have to install it at work too.
I have been suffering at the hands of MS IE of late. The three recent Windows versions 5.x, 5.5 and 6.x are quite different from each other, with a wide selection of defects and strange behaviours. The Mac version is totally different, having a different selection of defects, but is in many respects better than any of the Windows versions.
Basically all these bugs and problems, require so many hacks, and workarounds to make the browser display correctly that if it were not for their dominant market share, no sane designer would bother coding for them. To make matters worse some hacks have to be hidden so that they don't interfere with other browsers, or other versions of IE.
I was suggested the following article: "The IE Factor", which sums up my feelings and sentiments.
Due to overwhelming demand! well one request, I will make paxtv available for public use, soon. I just need to remove most of the scaffolding and legacy code, and write some nice documentation.
I've made a little tool to convert XML TV feeds into xhtml content. At first I planned to do it all dynamically at page request time, then cache the results using the excellent Cache::Cache module. The I gave up and decided to just grab the pages with Wget - on a cron job, and build the pages asynchronously.
It's a very simple application, load the XML data feeds off disk, run them through an XSLT engine, and stuff the contents into a web page, the four BBC channels on one page, and five commercial ones on another. Dead simple, now I don't have to use the awful EPG that comes with FreeView or buy a TV listings magazine.
For the last few days I've been working on our new web site at work. It's been a mix of xhtml, css, and dhtml, and so far it's been quite fun. Part of the work involves generating 500 web pages from XSLT, which is where the pain began.
Turing a page of XML into xhtml is quite easy, as long as the result xhtml you are aiming for is reasonably sane. The xhtml I had came from Dreamweaver, so it was quite insane. I decided it would be easier to re-write from scratch, which is what I did.
An hour or so later, I had a nice XSLT sheet, that spat out a nice xhtml page, the CSS worked on it, and it looked like the xhtml that the graphics bod had originally created. As expected it worked perfectly in Mozilla Firefox and Opera, but IE was quite broken. It's taken me two days to fix all the problems that IE has, and it's been very frustrating.
On the up-side I discovered an excellent CSS web site /* Position Is Everything */ and I know a lot more about xhtml/css than I did at the start of the week!
Our village has a web site. It's great that someone did something and made the site, however it's of such a poor quality, that one does wonder if it was worth the effort.
It's a classic Microsoft FrontPage monstrosity. It's got "one of everything" slapped down all over the place, with no overall feeling of cohesion. The HTML is awful, CSS is a joke, nothing validates and it's an accessibility nightmare.
I wouldn't mind, except some of the budgets is coming out of my local taxation.
It was snowing on our way home today. By the time we had walked from the station to the front door it had pretty much stopped, and you would be hard pressed to have found any evidence that it had ever snowed.
Tomorrow it could be the top headline in the regional news....
We are working on a new site at work to replace the current one. I bitched so much about the current one, that this time the graphic artist has used more of his skill than Dreamwevers crude automated code. The result is a base template that is more than 10KiB smaller, and each content page is also smaller. Best of all it looks nicer too.
For my part I've worked on optimising some of the ECMA Script and the style sheets. This time he understands my Perl based templating system better, so has been he has been able to use it more to his advantage.
We shall see if we go live on or about the right date, all I have to do now is such a shed load of data out of the SAP system, and template up about 500 pages of product data.
Last weekend I wrote a little backup script. Nothing fancy, it simple TARred up a bunch of folders, bzipped them nice and small and put them somewhere for burning to CD-R. I don't have the luxury of a big file server at home, so I've not copied the backups across the network to file server to keep.
While installing the mozCC tool
on my browser, I borked my
Bookmarks.html file. Thankfully I was
able to pop over to my backup directory and pull a nice copy out.
Well it turns out that one of the DIMMs I was sent was bad, and the Sun wouldn't use it. I re-seated it twice, and tried it in a different socket, but it was to no avail.
Still 224MiB is a lot more than 32MiB, so I'm not going to complain much!
So I've moved all my machines around and wriggled the SPARC free, opened it up, added the new DIMMs and booted to a happy open firmware state. Put it all back together and let it boot again, and then started the OS.
FATAL mistake, Solaris gets about 30 seconds into the boot sequence and panics with an "asynchronous memory fault" and stops. DIMM 1 is not happy, either it's not properly seated, or it's duff. It's too late to do anything about it tonight, but I can see a lot of fiddling about tomorrow!
I did it again, I made some minor change to my xhtml, and broke it. Annoyingly Mozilla carried on, and it took a bug report from a Safari user for me to see it. A quick validation check showed that I had some badly nested tags, now fixed, and I quickly checked with Konqueror to confirm that the problem was fixed.
I did some additional Konqueror 2.2 tests, and the navigation widget doesn't work - which is a pity. I don't have a later version or a Mac to test it with Safari, so I don't know if it's something that is now fixed in the khtml rendering engine.
A nice friend of a friend had just the right kind of Sun DIMMs for my ancient SPARC station 5. He put them in a box and posted them over to my friend, who gave them to me this morning. Even better is that with the US Dollar in free-fall at the moment, when we convert from dollars to Pounds Sterling, they come out quite reasonable, 25p per MiB.
Now all I have to do is remove the stuff from the poor little SPARC box, open him up, and plug the DIMMs in.
The UK's antiquated rail system has a terribly designed on-line web site that allows you to plan your rail journey. It's an access nightmare, awkward to use, and frankly an embarrassment to the Internet.
While hunting round the WebStandards site I spotted a link to a nice version of the site. It's not a cleaned up demo, but a fully working interface, using the same backed data. If you want to find out about train times in the UK I heartily recommend this site, instead of the official one!
In the past I've just hacked on Perl code while it sits on the file system. For a long time I've known I should use some form of revision control system, I've just never got round to setting one up, and understanding how it works.
Today I imported my modules into my CVS server, and started to read the man pages. Doesn't look to scary, once you read how it works. Anyhow I shall see how it goes....
A number of years ago I wrote a very basic Perl module to parse configuration files. It's something everyone does, and CPAN is full of them. I never released it to CPAN, as I saw little point.
As part of a project at work I decided to dust of this old module and update it, and this time it did go up to CPAN. At the same time I realised that I could patch one of the existing modules to give it extra functionality, similar to my requirements. So I release my module to CPAN, and sent a patch to the maintainer of the other module.
Time passes... It turns out that the patch, and indeed my original version suffer from a problem on Windows platforms. Windows and DOS systems use the back-slash symbol for their path separator, but I'm using it as a continuation symbol, as Bash does. Even though I use to use Windows, I always used the Unix path separator, slash, and let Perl do it's magic when it was accessing the file system.
Today has been a day for cleaning up directories. At work we are about to begin a cosmetic upgrade to our web site, so I've decided to try and enforce some order on our directory structure. At the same time I've swept away all sorts of file odd'n'ends to give us a clean canvas to work from.
Friday is a short day at work, so I've had enough time at home before dinner to clean up my home directory on my Linux box. When I upgraded from Windows to Linux I copied many files and directories over en masse without checking to see if they contain anything useful. Today I cleaned things up - or at least moved most of the ancient stuff into archive folders out of the way.
Last night I left wget running, pulling the latest Debian 3.0r2 SPARC ISO off a mirror. I just checked the md5sum, and tomorrow I'll burn it to CD in preparation for the upgrade of my SPARC station.
Now that I've ordered some DIMMs for my SPARC station 5, I've started the download of Debian Linux 3.0r2 for SPARC. It's only 660Mb for the first disk, and with luck I'll be able to pull any additional data of the net rather than downloading the other 5 disks.
It's a shame to be removing Solaris from this box, but Debian will run better on it given the machines limited hardware, and it will also be more familair to me, now that the rest of my systems are also running Debian.
For a while I have been tinkering with the idea of buying a new system. I don't want to spend a lot of money, and I don't want a huge box. Someone on use Perl; suggested a system buy LinITX as an ideal small base for a home system.
There are quite a few manufacturers of mini-ITX based systems out there, and most offer low-cost systems with no OS or Linux pre-installed. Most of these systems are based on a VIA Eden or C3 mini-ITX motherboard. Obviously they are not the fastest machines available today, which is okay, but it sounds like there are enough snags with Debian to put me of them for the moment.
A nice Micron employee called Tom Schmidt has written a Perl script (memconf), that parses the output of the Solaris prtconf command, and works out how much RAM you have installed in each DIMM slot.
My Sun box doesn't have enough disk space to actually have Perl installed, so I sent him the output of prtconf -pv and he worked it out for me. My system has a single 32Mb DIMM in, so I can now safely buy 7 second hand Sun DIMMs to take the system up to it's maximum of 256Mb.
Yesterday I tweaked the Samba settings on my Debian box. Mostly to add a SMB time server option, and to tidy up the configuration. Last thing before shutting down the Win98 notebook, I checked that everything was working fine. I even booted things up this morning and made sure the notebook was happy.
This evening when I got home the notebook as not happy, it had lost the network, including it's self, and was a very unhappy bunny. I tweaked with the settings on the Linux box and tried again, and all was well. We shall see what happens!
After more silly name changes Mozilla Phoenix/Firebird has become Firefox. Version 0.8 of this many named browser arrived this week. It's still a beta version, it could do with being smaller and faster, and it's still missing the odd feature in it's bigger brother, but it's a cracking browser however you look at it.
Between my work, and the train station home there is a small, and some what elderly Sainsbury's supermarket. I don't like it a lot, and the only reason we shop there is that it's convenient.
I can cope with it being small, I've been in smaller but better stocked stores before. I can cope with it being a tad shabby, it's basically clean, and that's what matters. Basically it's not competitive on price with Tesco or ASDA, and the quality isn't as good as Waitrose, or even Tesco to be honest.
What really annoys me though are the small things, like an in-store bakery that never seems to be able to bake decent bread. Or their 1p refund on shopping bags that you never seem to get. Their latest annoyance being their "scan and pack" policy, where they take an age to scan and badly pack your shopping in a collection of new bags!
Thankfully, a new Tesco Metro is opening along my route, Sainsbury's will either have to improve, or lose my business!
Unlike some people who boycott this global e-tailer because of issues regarding their patent claims, I'm boycotting them just because my last order took over a month to arrive, yet everything was on 24-hour availability. I can accept that anyone can have a bad day, but they never bothered to reply to my complaints - which is just bad form. So far I've gone three months without them, and in excess of £100 has gone to other retailers already.
Blosxom is a very simple to use, yet powerful blogging tool. If we had any news content, I think it would be the perfect tool to build a news site for a work.
The only problem with the tool, is that it's far to easy for someone to feed blosxom an invalid xhtml story - poisoning the site. One way to maintain the content would be to run a cron job to scan each story, and to make sure the content is well-formed xhtml. The XML parser in libxml2 is a very good candidate for the job.
My computers at home run Unix of some flavour. My better halves computer does not, it's an old Dell notebook running Windows 98.
Without running Samba on the Unix boxen, the Windows machine would be cut off from the printer, and backup via CD-R. This would be bad, and would be a serious impediment to my gradual conversion of all my computers to Debian Linux.
Anyhow Samba is now correctly configured, and files are again crossing the switch and domestic IT harmony is restored.
Eventually I'll have to upgrade the notebook, but I think that getting some form of Mac will be the best path to Unix utopia. I'll then have the old Dell, and try and put Debian on that for myself.
If you ever want a quick, simple and modular wiki, then Brian's Kwiki is worth looking at. It's dead easy to install, and while not as feature complete as others, it offers a good selection of features to cover most needs.
I like it so much that I've contributed to the xhtml.
Today I had my weight measured. I'm still over weight, but it has come down 4Kg since last year, so things are moving in the right direction. My target was 85Kg, which given that I have a heavy broad frame sounds like a fair target.
For fun I calculated my BMI, it's too high. I then re-calculated with my target weight, it's still too high. According to the US CDC I should weigh between 53-72Kg. If I did weight 53Kg then I would look very thing and sick. If I got down to 72Kg then would still look pretty thin, and unwell, but not emaciated.
I know it's a simple formula BMI = weight (kg) / square of height (m), but if you ask me, volume is a cube relationship with length, not square. So BMI values can't be very meaningful if you are tall or short. Didn't the medics who came up with this do any maths at school?
There are several browsers in the world, and they all do things different. If you do things properly, things work predictably on Mozilla and Opera, and IE may or may not behave. I don't have a recent khtml browser, but I gather it's pretty good too.
When you make a small mistake, it can be a real pain locating it, because one browser doesn't show it, and the others express your error in conflicting and unhelpful ways.
I've fixed one more little defect here, this time a tiny style sheet glitch, that Mozilla was mostly okay with, IE did odd things with, and Opera threw a fit - thankfully in a diagnostic manner.
The more I think about it, the more I realise that to use the Unix GUI effectively, you must first understand Unix. Otherwise you end up miss-using the tool and wondering why it doesn't do what you want.
When I first came to Unix (Solaris) I had no idea what was going on, it was very much different from my previous experience with DOS and Windows 3.1. I started to use the GUI, but found it awkward and unhelpful.
I now use Debian/KDE for a desktop system, and while the GUI has greatly improved, I think I find the system easier to use because I understand the shell better, rather that I use the GUI. Most of the time, I simply open up a console window and work at the command line..
I've been invited into Orkurt. Doesn't seem to take long to get invited, and quickly aquire "friends". It's a bit of a cheat really, as the only friends I currently have, I knew already!
Today I have been wrestling with Dynamic xHTML (DHTML). The xhtml side isn't too bad, Mozilla and Opera are well behaved, and if you set the correct DTD, even IE6 behaves it's self most of the time.
On the CSS side, IE is again defective, but at least it's something I'm fully aware of and I know where to look to find fixes.
It's on the ECMA-262 and DOM that I have problems. IE just seems so buggy and off the wall, combined with my lack of familiarity of these two technologies, makes this bit very hard going.
However, as much as I hate IE, I think I will overcome it's problems, and produce something standard compliant, and working on the main browsers.
I've been offered some RAM for my old Sun box. I know it has 32Mb installed, but I don't know if that's 4x8Mb or 1x32Mb. Because the box is sitting below other boxes, and I can't be arsed to move them all at this time of night, I can't just have a peek inside.
So I thought I'd just poke about as root, and get the Open-Firmware to tell me. Snag is I can't remember what the password for root is. I feel a tad bit stupid at the moment.
As predicted by Netcraft SCO removed their DNS entry and let their domain go of line while the MyDoom DDOS storm blows over. Obviously they don't need their site that much, since they became a legal parasite, so they can afford to sit it out.
Shortly the next MyDoom virus will attack Microsoft, who do have to keep their web site up and running. With a few billion in the bank they can throw money at their problem while the storm blows, I wonder if they will revert to using Linux again...?
This weekend I have been playing with making "drop down" menus with DHTML. I've cut away at the code as much as possible to make things easy to see and maintain, but I'm still not happy. The DOM methods in ECMA-262 are okay, but not perfect, and though ECMA-262 isn't bad, I don't like it as much as Perl.
I've got this blosxom thing sorted out and working more or less now. Permalinks seem to work okay, and I've got the style sheet and templates sorted out. It's simple enough to use, and seems to work okay, though I'm not 100% happy with it yet. I'll play with it for a while, before I decide if I should upgrade to something more powerful, or stay put.