Results tagged “Corporate”

Finally I got a wired LAN access instead of this crappy WLAN connection. Now it works like a breeze and I even have access to a broader range of machines.

On the workside things are going around a bit without control. Two people are on vacation and the work concentrates on a few bottleneck employees. One of them is my group leader. I told him that I think that he has now too much work loaded on him but he doesn't think so... well.

I also have a hard time waiting for a new Wikipedia database dump. At this moment the Wikipedia and Wikimedia pages are down for "hardware maintenance". I hope this also involves a DB dump sooner or later...

| | Comments (3)

Today I stumbled over an error message when I tried to visit a site with IE. Suddenly a error message came up instead of the site. "Company IT security policy blocked access to this site. See blah, blah, blah...".

A link directs to a notice which explains why there is a block. "Not all traffic on the * network is business related." You tell me. "Some of this traffic is less than professional and [not acceptable]." Hell yeah, but you forgot to block "The extend of this traffic is not great, however, I'm certain you agree [anything is too much]." Surely because we need bandwith to connect Outlook to our central US mailservers and have all mails travel around the world every time I want to show them on the screen. And if I have to wait only two-and-a-half minutes instead of three when I just want to have a look at my Inbox this is a leap forward even if it takes another coffee-break if I click on a mail to actually display it.

So the companies clever IT heads have decided that I must not visit certail sites using their proxy. Well, I'm not using their proxy for that non-business stuff because I use Firefox and our local internet access for that normally. Just had to use IE for a certain MediaPlayer-enabled site showing a dancing Citroen robot.

I guess they will get rid of such backdoors somewhere in the future...


Wow, what a weekend. Yeah, I know that it's Thursday but I've been busy since last Friday in a row.

The weekend was filled with an update of our software platform to the very latest version, involving changes at almost every place in the system. There were also changes necessary in attached applications which are completely out of our responsibility, but because no one is around anymore we got the job to fix these too.

All in all we started at Friday evening with database preparations needed for the software switch between Saturday and Sunday night. The database changes went almost flawlessly.

Between Saturday and Sunday we changed the software part of our platform too which wasn't that flawless and had some very unexpected occourences but finally we managed to bring the system up almost in time. During the whole process, customers should have noticed problems only three times for a few moments. We're getting better every time :)

Sunday to Tuesday we were busy almost all day long fixing incoming bug reports and minor changes but by Wednesday the System was working as expected.

We'll still be busy with some afterwork for some time but that'll have an end.

The client seems to be quite happy with our work and I guess that there are already other projects lurking around there. There is a chance that this client will grow quite a bit and in that case we'll be there to feed it to make it big.

On the other side, we're all quite exhausted from that weekend but that'll go by too. And I think we're all proud of our common success.


I don't know why, but today I gave Trillian another try.

I tried it several weeks ago but deinstalled it again because I couldn't get it to work reliably through our NAT and firewall. Either it didn't connect or if it connected, the ICQ-server didn't accept my login.

I would be fine with ICQ itself but some of our customers use Yahoo Messenger to quickly communicate with us. And in the end the memory footprint of both, ICQ and YM, was just too much to get comfortable with it. And I decided to switch to an IM which supported both. Well, that was some weeks ago and was a complete failure.

But this time everything worked seamlessly. I installed Trillian, gave it my logins and pwds and set the connection-setting to some un-blockable ports like 80 (HTTP) for AIM and 443 (SSL) for ICQ. Connected and whoa, everything green. No error messages and all contacts present.

Its your luck Trillian that I give faulty software a regular try-again.


Last week I moved from the workplace where I worked for the last three years to a new workplace in another block of the building.

The reason of that is because another project caused the other people of my team to move out several months ago. I were the only one left. Didn't matter that much because I have responsibility of completely seperate stuff.

But now I finally decided to change my location too because the people of the other project distracted me too much from my own work.

The move took around six hours split over two days and now I've ended up with the largest workplace of the people here and enough room for all my small and large stuff. I took everything with me, computers, papers, equipment, pictures and even plants I've grown and acquired from others over the time. Well, the old place looks quite sad now without the slightest green but that's not my problem anymore. I just left my old chair behind because a new and probably better one was standing around lonely at the new place.

The only things which got worse are the larger distance to my teamleader and the worse (if even existing) WLAN-connection.


Trying the Google Desktop Search works flawlessly on all machines except one. It's the one, where the Wikipedia dump also resides on and the errors are those: * indexing in background stops after ~15 minutes and doesn't continue until I restart the GDS * it completely left out the directory with the Wikipedia dump initially

But now, after I had forced the initial index process to complete by always restarting the GDS, it seems to catch up all forgotten stuff. With Sysinternals Process Explorer I can see, that the GDS is walking through the Wikipedia directory...

I somehow suspect that Scansoft Omnipage Search Indexer plugin to be the root of that evil, because GDS always ran flawlessly until Omnipage has processed several files.

Moreover, Scansoft seems to have changed its registration policy: The first computer I installed its plugin on needed to register, which was done on a single HTML-Form and worked without further notifications. But on the second computer, where I installed the plugin two days later, the registration process spanned several pages and ended with a note, that the plugin is beta and would work until April 15th. Great.

On the other hand I had some troubles bringing my personal laptop online with WLAN and VPN at the company. I had a working WLAN connection but the Cisco VPN client always refused to connect.

After three days of trying now I succeeded by deinstalling local firewalls and virus scanners completely, updating the WLAN drivers and downloading an updated version of Ciscos VPN client somewhere on the net. Normally only customers of Cisco can download it, but someone at some university forgot to secure the internal VPN-download page ;)

And after all of that hassle an email arrives, that VPN will be restructured next week to achieve SOX-compliance...


Seems, that our company is going to include some passages into our work contracts, which in my opinion are already given by law in the collective contracts for IT professionals.

But until now, if asked, the HR department everytime replied with something like "this doesn't apply to us because this and that". Well, if I understand the collective contract correctly (and it isn't that hard to read), everything applies to us OF COURSE! But in the past I wasn't that unhappy with their denial, that I had to insist in my rights.

If everything is now written down one more, I have no problem with it that our HR can't say "no" anymore :)


Today I accidentially did something which easily could have cut my head in company.

Tomorrow we have a release at a customers place, where we bring up an additional machine to handle the increased load.

I was creating configurations for the new machine in the administration interface when it failed and just returned an error. I was forced to directly connect to the database to create the entries. During the creation I stumbled over some small errors and decided to throw away the whole table data and copy everything from the master database when I realized that I actually was working ON THE MASTER DATABASE! I had thrown away the master data, dammit. Adrenaline... Sweat... Has someone seen that? No? Good. Ok, hopefully none of the running machines restarts to fetch its now vanished config. What do I do now? Re-fetch the data from one of the client databases. Dammit, replication has already deleted the data there too.

More sweat, more adrenaline...

But wait, what was one of those minor errors above? Whoa, replication to one machine was failing, the data could still be there... Immediately connect to that machine, stop the replication mechanism, check for the data. YEAH! Still here, whew.

Copying back the data from that failing machine was easy and luckily no one noticed that incident.

I have to be much more careful in the future, when I deal with databases. Have to get practice on that...


There is a lot of work to be done in the next few weeks. Aside from minor updates at several customers one customer wants a major update of the whole software to the almost latest version. Currently he runs a version which is almost three years old.

For us this means that we have to create a update-scenario where everything from DB over backend straight to all frontends have to be replaced with their latest counterparts.

This woulnd't be too difficult if we wouldn't have to take care that if we don't handle the update we can safely step back to the previous version without much impact.

There are also further smaller releases and other deliveries in other projects to plan and to deliver which increase the overall worload-level quite a bit.

I hope, we'll handle that until mid march.


The last week was a bit stressy for me. Besides a shift in my work area, I've been busy organizing the Lord of the Rings Extended Edition Triple Feature.

I managed to acquire a conference room in our companys building for that day.

The feature will start at 11am and will last until midnight, only interrupted by two 30 minute long breaks between the movies.

Today I want to test our equipment in the room to find out if my Creative Inspire 5.1 Digital 5700 is strong enough to provide a propper sound environment in that room. I also managed to borrow a 16:9 video projector.

On the other news I switched from C/C++ development to web development using PHP at work. I did that because our group is in critical lack of web developers and we already have quite a backlog of stuff to do in the web environment.

The current web source is a pile of mess. I'm trying to refactor some pages and introduce a template system (Smarty) to take the presentation out of the business logic in the pages which currently is non-existent. It's quite hard because the current system has many includes on each page and almost everything is dependend on something else and at almost any place in almost any include there could be output written which you can't avoid.

CSS styles and JavaScript are included on every page and I tried to split them off into external .css- and .js-files but have been only somewhat successful because even those scripts contain dynamic content which could change on every page. Dammit.

I'm still unable to give an estimation to my group leader when or even IF I'll be finished but I still have some motivation left to spend on that whole stuff.


Two days ago we had a release of our product in the middle of the night.

Initially had forgotten to update something in the hurry which then fell back to us and two of us had to stay here and redo half of the stuff until 3am. I`d been one of those two.

But today I`m almost back at default productivity :)


Last Thursday I finally finished a feature in our product which I've been working on for two weeks.

It was a bit frustrating because the automatic testcases for the product had to be corrected for the new situation but when I fixed the testdata for one case another test freaked out. Back and forward all the time; change data, compile, run 390+ tests (~12 minutes), evaluate error messages and back to the beginning...

But finally I finished it somehow and for my own surprise I only had to adapt the testcases; my implementation of the feature had worked all the time.

Next thing on friday was that I implemented another feature, which I approximated to need five days to develop, full featured (including testcases) before noon. Four hours instead five days. I was very surprised again when I found out that it worked and nothing was left to do for it.

These two things increased my personal motivation quite a bit.

Another thing last week: I saw that several of our quality assurance people had a long talk with our development manager. I have no idea what exactly was talked about but I have a good feeling because recently I often talked to them about our development processes, structure and quality measures. We all quite agree that the current structure and processes are not suitable for development. Our alignment on projects instead of products constrains us more than it aids us.

Along with that thoughts I found a project called GForge which is like the Sourceforge environment. GForge is based on an earlier CVS snapshot of Sourceforge but has been development independly since that time. I thought about if this could assist us in our development and played around a bit with a test-install of GForge, but I'm still unsure... Have to talk to our Quality Assurance people about that. It surely would be neccessary to integrate our installations of TaskZilla and TWiki with it.


Sorry for not posting immediately butattending OS04 took until late in the evening and after that I didn't think about blogging.

The OS04 was quite interesting and one of the best things of all was the keynote from Jon "maddog" Hall. In general he compared programming with creating art, how he came to programming free software, how he came in contact with Linux and lots of small anectodes from his life. The fact that his speech took three times as long as planned (90 instead of 30 minutes) and nobody really cared about that speaks for itself.

Later I watched the presentation of the Irrlicht Engine, an open source realtime 3D engine written in C++. The real cool stuff about it is that it is possible to open and display a Quake III level and be able to float through it in roughly 5 lines of code. Whew...

Next point was "Soft Migration" where a nice lady spoke about how to migrate business applications used in companies to open source alternatives. She had good points about the migration but later she got lost trying to explain why it is bad when a state founded "knowledge center" tries to help the open source community and that it is good when a private company would do exactly the same. For me it seemed that she had a simple impression from how the economy and community interactions work. Almost too idealistic...

Moagg was next on my list, a cave-flyer game project created and programmed by one of my work colleagues. Looks fun!

In the next unit we hear about what to take care of when you replace your Windows NT Primary Domain Controller (PDC) with a Samba server. Easiest way works with three clicks in Suse Linux.

Staying in the same room after that talk brought us to "Quality criteria in software projects" by Dr. Klaus Schmaranz. He talked about a large scale EU-pushed project which failed not because of technical but because mainly of management incompetencies. This failure could have been prevented when the management would have recognized some well-known software project guidelines which are already existing almost thirty years...

The talk almost seamlessly continued with a presentation of Extreme Programming and its pros and cons from the view of the referee. Was followed by an active discussion of how extreme programming could be applied in common corporate environments. The referee was quite surprised by the feedback which he isn't getting from his studends normally. Well, thats the difference between theory and practice ;)

This last talk ended at about 19:00 and altough there were still points on the list but I decided to skip them...

All in all it was very interesting event, and I'll look forward attending the OS05 next year.

I also found other impressions from the OS04 on the net: Alfie's OS04 report


Today my mail computer was changed. Instead of a 800MHz/256MB AMD K6 machine I now have a 2.8GHz/512MB P4 which enables me to wait for the global Exchange server ten times as fast as before. Yeah!

But my development machine is still unchanged at 800MHz/384MB AMD K6.

Since I'm not permitted to switch the purposes of both machines but only to boot a second OS on the Mail machine (loosing my ability to read mail) . I found following half-legal solution: 1.) grab one of our VMWare packages and install it on the Dev-Linux on the mail machine 2.) boot the corporate image inside the VMWare

This should enable me to read mails AND develop software on the high-performance machine.

Well, but first I have to install Linux on it....


We have some trouble with a customer not paying its bills for a project.

So our management had the idea to introduce some limitations during a maintenance update of another project with that customer without telling him what we are up to.

I don't think that this is a good idea and can make the customer quite upset if he realizes the next morning after the update that in the logfiles lines like "Limitation reached" are appearing and disturbing their operations.

This is certainly not good for our companies reputation.

There has to be another way but who in the management level wants to listen to a simple developer...


stretch... Yep, back at work. And almost anything is left as I leaved it behind.

No customer action, no emergency, no progress in my immediate area. So no notable stress-level.

Just an hour ago a meeting ended where our head of overall development picked together people from various (independend) teams and discussed problems on a specific project. This project meanwhile has quite a log of defects and errors. So we discussed the top defects in a medium group.

All people brought in their experience with similar problems on their own projects and some different possible solutions and views to the problems came up and I personally think it has been productive for everyone. It was also a knowledge-exchange for the people not concerned with this special project.

If you have long-lasting or always recurring problems on your project, try to discuss them on a slightly higher niveau with someone who has nothing to do with it. But be careful to pick someone with experience!


Somewhere on a page of Joel Spolsky or one linking to it I stumbled across the comment that if you want to improve your software development but have to fear refusal from your colleagues you should try some creeping methods.

Today I tried that and "forgot" a copy of The Joel Test at a prominent place, right next to the coffee machine. Perhaps some people will take a look when they are recharging their energy with caffeine and unconsciously pick up a meme from it.

In the long term my goal is to have an influence in our process of software development so that we can enhance our overall quality and maintainability. Perhaps I can even get my foot into the newly re-founded Software Quality Team, currently only represented by one person.
I have so many ideas, some of them surely worth a deeper evaluation:
- enhance/update our task management system, so that we can more easily keep up with Bugzillas development
- set up a better automated build machine, better integration with CVS (Tinderbox?)
- enhance our make-scripts for faster from-scratch-building
- use php templates in our web administration interface for easier customization and maintenance
- use Eclipse IDE for most of our development tasks (C++, PHP, etc.)
- ...
Gaaahh, my head is bursting of dozens of small ideas how to improve development. Just have to convince the right people...


Last friday I was finally forced to install Yahoo Messenger because some of our customers use this as the main comm-channel to our staff and I don't think that they'll install ICQ just to be able to talk to me.
Customer rules, that's how it should be ;)

Hopefully we get the tests done today. If the customer wasn't that "lazy", we could have done them two weeks ago.


I almost forgot to blog recently.
This could be caused because in the last week I could program and develop freely most of the time.
There were no tasks in queue which could be worked on without help of any of my colleagues which are/were on vacation and cutomer requests/reactions were low-water too.

I worked on many parts of our codebase and threw out many leftovers from already refactored structures and updated the code which was still using it. In the same pull I also refactored a large suite of testcases and simplified >5.000 lines of code down to about a sixth of its size.

Some of my refactoring activity also affected performance more than I expected: I managed to speed up some often-used parts near the DB backend to be about 40 percent faster than before. And that in an area which I expected to have finished performance optimization sometime last year. Yay!

I also managed to get my ICQ online again with help of the ICQ Support Team. The problem is/was that the server lying behind port 80 of seems to have a problem with my ICQ account. ICQ Support suggested to switch to port 443 (luckily not blocked by our NAT) and everything worked again.

Within company I'm trying to get some abonnements and books ordered which I find quite helpful for developming as well as project management and software/project engineering in general. But I got no response yet from the person which is responsible here.


The last two days I had to fiddle around in DBs and logfiles to create and/or confirm statistical numbers a customer sent us from our system.

The problem was that according to his calculations we were loosing nearly one percent of all billing data.

He collected the numbers by counting through our logfiles and internal DB tables with grep and custom SQL selects.

Of course he got his numbers wrong.

But it was on our side to prove, that he was wrong and what exactly our numbers and his numbers contained.

Of course again I got it wrong too the first time, because our specialists for these areas are on vacation and I had to dig in into the DB structure without exact knowledge of different columns meanings.

Well now, after two days and with the help of another DB specialist we could send the customer a (hopefully well enough documented) explaination of the numbers and how we acquired them.

Never let a customer dig around in your internal structures!


  1 2 3 4 5 6 7