Edward looking helpful

"I have created this website with the help of Joanna in order to give me something to do while I am waiting to catch my flight to Beijing. I can't guarantee that any of it is actually true or not, but it might be slightly helpful or amusing to look at if you are interested. If you have any constructive suggestions, you can always try Edward or Joanna via e-mail."




* your suggestions are very helpful, thank you

Site History:

Creation date.

Provisional Itinerary and map produced (i.e. I made them up)

What was more like a working itinerary supplied by Joanna. It covers more than 2k miles, but this is because of extra options.

Worked out the whole IE5 CSS <STYLE> sheets thing (i.e. links go crazee when one puts a pointer over them, see the itinerary page). Looks like I shall be creating a PERL script to convert my text file drivel about places I visit into itinerary pages as I move through china. This saves me doing it now, when all I have is a Lonely Planet and a set of Monkey books. Probably change Itinerary to something more normal sounding like, `Journeaux'.

Laboured away for most of the two days to produce a set of PERL and BASh scripts that do the following:

(hermes filtering) Organises messages from my hotmail adress into a mail spool (big file containing all the messages stuck together) on the Hermes message store. The penultimate and ultimate lines of each message are a date and place name respectively.
spoolFTP This BASh script retrieves the mail spool containing these messages
spool2txt This PERL script strips off extraneous headers/footers from each message and creates a file for each one based on where and when it's content refers to. I did it like this to save each trip report in text format for ease of long distance editting via the Caius java SSH client login.
txt2HTML This PERL script takes a set of SOMEWHERE??.txt files, where ?? is a coded form of the date of that place's report, and joins them up to make the place report.
txt2INDEX This PERL script takes each text file and extracts the content from it, and passes each place name to txt2HTML, while itself building an index page.
mainScript A BASh script to run the whole lot in order and generally try to stop things messing up.
(crontab) The door.acad.cai server runs mainScript 8 and 38 mins past the hour, and pipes it's output into the mainScript.log file. Any errors crontab encounters during the execution of the mainScript and it's children are locally emailed to me by Cron, and appear in /var/spool/mail/egs21 (it there is no mail then this file will not exist and you will get a 403).
localSpoolTidy Just takes the local spool, discards 99% of the headers, and makes HTML out of all the e-mails sent to me from nice old Mr.Crond.


  1. Polishing of crontab and mainScript to actually make it run, and addition of a few extra commands to help manual running. Tried sending a few extra e-mails and it appears to work.
  2. Also wrote a new script to tidy up the local mail spool a bit so that is looks easier on the eye up website way.

  3. Noticed a `feature' as well, if posts are made with the same date, use goes in order of alphabetical preference, followed by prioritising to newer reports. This is not a `bug' repeat not a `bug'.

  4. Also wrote a PERL script that converted txt into HTML by replacing \n with <br> and blankLines with <p>. This makes the links in the table above look nicer.

Read my messages from theCounter.com.

Completed a script (written over a 1500ms/8,000 mile SSH link to Cantab from Chengdu, PRC) that extracts some information from the wc command (wordcount) and tacks it onto the end of the itinerary index. It is a hideous hack: I open up the HTML tags and add more stuff then finally close them for good (see source), but it appears to work.

Stopped the crontab, thus the website is no longer active. THE END

© 2000 Edward G. Speyer & Joanna M. James.
Web space provided by Gonville and Caius College Student Web Server