Entries Tagged 'Uncategorized' ↓

The 3rd BCS CMSG Conference is now open for registrations with a good lineup of speakers, and an excellent looking venue.

A number of people commented on problems getting to Cambridge (not so convenient from various parts of the UK), so hopefully Oxford will fit the bill a little better. In addition, we can up the budget a little for a slightly more commercial venue.

Some other differences this time around:

• PB Projects is doing the bulk of the heavy lifting in terms of organising details etc – yes it costs us, but it seems to be working fairly well. and takes some (though not all – e.g. sorting out timetable/speakers) of the load off the committee.
• Only just realised that offering a group discount might help bring in a few more people – we more interested in having more people there than making a vast profit (haven’t yet managed to get the committee fact finding trip to the Caribbean off the ground to spend our current surplus, but am working on it…)

So, the first registrations are already in, and look forward to plenty more!

Meanwhile, am hopefully getting a little breather to allow rather more blogging than of late – have a stack of topics piled up awaiting that spare moment!

I went to the Serena UK Customer day recently where there was a brief discussion as to the reasons for the $1.2bn Serena buyout by Silver Lake Partners. The FAQ (follow previous link) mentions:

“We expect that under private ownership we will have far greater flexibility to focus on
meeting your needs more effectively and efficiently. Being a private company should
also allow us to make long-term investments in our product and service offerings and
ultimately become a more valuable partner to you.”

The presentation mentioned these investments which might depress the share price if it remained public. Not explicitly mentioned, were costs and other requirements of Sarbanes Oxley compliance. Interesting how Serena as a vendor of tools to help other organisations achieve compliance is itself avoiding the issue by going private!

There have been a few blogging delays due to such family things as a new son! But also changing hosts – everything now hosted at Textdrive.

As ever, migrations always take time and indeed the blog got left on the old ISP for quite a while as I couldn’t face moving it! It has taken some fiddling to get everything working as it should. In the meantime I failed to get to grips with the new setup for some weeks and committed the cardinal sin of breaking old URLs to blog entries – for which I can only apologise and take my medicine like a man!

Have been playing with Subversion as installed here as the new mechanism of getting stuff up to the Textdrive machine since straight ftp is outlawed and my previous mechanism involved a hack of KFTPUpdater – a Python script which only uploaded changed files, but did it via vanilla FTP. I must admit it is nice to have Subversion installed and available for versioning as part of my account. That said I had a few niggles getting it actually working – finally found the relevant bits of information and made it. (p.s. I don’t find Textdrive’s Webmin admin interface the easiest thing since sliced bread). Even better I now have Perforce available to me as well, and have a machine running as the Textdrive guys nicely gave me a port to use. Hassles there too as originally the binaries for FreeBSD4.1 core dumped when run on FreeBSD5.4 so I put it aside for a bit. However, some server upgrade later and they magically now work fine. Thanks to Gordon Tetlow I even have P4Ruby installed locally since I couldn’t compile against the Perforce API built with GCC 2.95 when Textdrive have GCC 3.4 installed (and I didn’t fancy faffing around with a local VMWare install just to get things fixed).

I have been doing some research into new blogging engines and rather like the look of Typo (and need an excuse to get to grips with Rails!), so am hoping to head in that direction. However, I will first need to tweak it to add some article versioning – am still amazed that people gaily chuck all this text into databases and don’t version it in any form… (Have heard regular horror stories of databases being wiped and stuff being lost through injudicious SQL statements – no thanks). Given that I now have P4Ruby installed here, I am rather hoping that it will be very easy to make a few mods and save versions in Perforce as articles are added/edited etc. Will post on my experiences.

The BCS CMSG event on 12th October on “SCM Agile vs. Traditional” that I organised went very well. There was a nice group of presentations and they seemed to mesh well and got very good feedback from the attendees. I look forward to a good set of talks at our conference next year as I think there is plenty of information to mine in this area.

A good proportion of the delegates (maybe 30-40%) had agile projects on the go already. A similar number thought they would soon have agile projects on the go, so it was good to see we had the right sort of people there. The Q&A session with the panel of speakers was interesting. A few people remarked that agile seeming like a fad. Also comments were passed on whether the government would ever consider more agile procurement instead of the current monolithic requirements exercises which end up with massive systems and often massive failures.

I took on a little too much myself in both organising the event, getting speakers set and writing a presentation myself, although Brad provided some excellent input, and indeed the basis for many of the slides via email. The presentation was fine, though I felt I needed to streamline it even more – still too much stuff!

As it happens, I presented pretty much the same thing at the Perforce User Group meeting near Cambridge on the Thursday and it went a bit better second time round. I was a little surprised that fewer people proportionally were doing anything agile like.

It was a great pleasure to collaborate with Brad Appleton on the article that was published in CM Crossroads in the August 2004 newsletter that covered releasing.

The article grew out of the talk I did for the BCS CMSG group in June. Perhaps as a result of practice, I felt the that article worked a little better than the presentation (which was a little too generic). Of course, Brad’s suggestions were helpful and improved the article, but perhaps a lesson is to write a paper first and then develop the presentation from that. We’ll see.

Anyway, the collaboration with Brad looks like it will continue, with a new presentation for the BCS CMSG event on 12th October on Agile vs. Traditional SCM. Something to keep me on my toes, and I look forward to continuing it!

It’s taken a little fiddling, but I have tweaked Rublog to do the basics for me in running this blog.

The main things I was looking for in blogging software were:

• Local editing and uploading of entries as files (to allow me to version control all entries locally before upload)
• WYSIWYG editing of html files

Links with SCM or version control tools seem very limited with most tools which I find very surprising. I don’t want all my content just sitting on someone else’s server with all the attendant backup problems. Also, I don’t fancy a database for blog entries as they are very difficult to version control.

I did wonder if I could tweak Rublog to generate static HTML which I could then upload, but realised that things like the calendar worked against that. I also found an ISP (beyondperception.net) who has Ruby installed and who has allowed me lots of freedom in installing other packages and configuring things.

It would be nice to have a commenting capability in Rublog, and who knows, I may get around to a simple form of that.

I was testing with Apache on Windows and then uploading to a Unix server. As a result, I wanted to create a directory structure which allowed easy copying of files around. I also discovered my ISP had only barebones Ruby installed with no libraries, so I had to install those appropriately. Adding to $LOAD_PATH and using relative pathnames for my data files fixed that.

One extra change was to add a function to read the mtime (mod time) of the blog entries from an RCS keyword ($Date: 2004/04/09 $) which is in HTML commend header. This is auto updated by my version control software (Perforce) when I submit the file, rather than when it might get copied up to the server which could be sometime later.

Mod_rewrite fun

My installation was made a little more interesting by the fact that I needed to use Apache mod_rewrite on the domain name for this blog (pointing it at a sub-directory of the main site), and I wanted to have nice clean browser URLs. Thus URLs of the form http://mainhost.com/subsite/blog/index.cgi should be converted to http://subhost.com/blog, and such changes should also be reflected in the URLs used within Rublog (e.g. for links and print styles etc).

PragDave kindly pointed out that he achieved the clean URL effect using a single ScriptAlias in his Apache configuration with no changes to Rublog. Unfortunately this wouldn’t work for me since I use shared hosting and only have .htaccess files to use to configure this sort of thing (in which ScriptAlias is not allowed). Thus I got to mess a little with mod_rewrite, and I also needed to change the Rublog code a little to make URL creation happen in a single place and tweak that function (instead of just reading ENV['SCRIPT_NAME']).

What I ended up with in my .htaccess seems pretty simple (just took some trial and error and local testing so that I could get at the mod_rewrite log to see why things weren’t working along the way):

# Redirect the robertcowham.com to sub-directory and blog references to appropriate .cgi
RewriteEngine on
RewriteBase /

RewriteCond %{HTTP_HOST}  robertcowham.com$
RewriteRule ^blog/$       /robertcowham/blog/index.cgi  [L]

RewriteCond %{HTTP_HOST}  robertcowham.com$
RewriteRule ^blog(.*)     /robertcowham/blog/index.cgi$1  [L]

# Anything not already caught by the above rules will be caught by this next one.
RewriteCond %{HTTP_HOST} robertcowham.com$
RewriteRule ^(.*)        /robertcowham/$1

With a single .htaccess in the robertcowham subdirectory which turned “RewriteEngine off”.