Phine Solutions web work notes

I have a few computers (desktop, laptops) at home and each one of them accumulate data and media files over the years. I used to copy them to CDs for backup purpose but this method has become too tedious. I do have a portal hard drive but carry it to each computer and copy the files is just too much work for lazy me. Lately I started to explore the options like a network drive and after some research this Bufflo Link Station network drive seems to be the best one on the market.

I made the purchase from buy.com and setup was very easy. Here are some points I want to share:

I plugged the drive into the network and power it up, and almost immediately I found it from my router’s admin console. So this is pretty much all the initial setup needed.

I went to the Link Station’s web admin console using the ip address that my router assigned to it. I went to the network menu and change the DHCP to static IP address, and changed the gateway and DNS server to the ip of my router. In this way, the LinkStation drive has a static ip.

From each computer, I mapped the network drive to the ip of the LinkStation. And there I had it, the network backup that I have been longing for.

I didn’t install the Memeo backup software that comes with the LinkStation since I use this one for my backups.

• The speed is good enough for the backup.
• The user management and trashbox feature is awesome.
• The drive only makes a little harddrive spinning noise when there is active file copying going on. I placed it on my book shelf in the living room and it is very quiet.
• There are some other nice features such as emailing and FTP access for advanced usages.

Overall this is a great drive.

I have an old Dell Inspiron 1000 laptop. It is slow (with XP), noisy, but it has the best keyboards in all my current or previous 4 laptops. I installed Ubuntu on it this weekend.

Actually it was Xubuntu, Ubuntu’s little brother. Since it uses Xfce desktop, it uses less resource and is more suitable for machines with less horse power.

Installation is quick and simple. Just boot the computer from CD (burned with Xubuntu image) and you will have a Linux desktop. Run the installation and it was quickly done.

Ubuntu claims “just works”.

This is largely true. I have a Logitech wireless mouse, and it was recognized immediately started working immediately. The network card worked fine too.

The problem was with the wireless. I have a Dell wireless adaptor card and it just wouldn’t work. There are quite a few disussions and intructions on the internet about how to setup or configure this type of card but none of them worked for me. It appears this kind of card was made by “Broadcom” and it is notoriously lack of Linux driver support.

Frustrated I took my Belkin USB wirelsee adaptor and plugged it in the Ubuntu machine, and Woohoo my wireless is ON. Not knowing how many hours I have to spend to get the other adaptor to work I decided to put my personal ego down for now and stick with the USB wireless.

Besides the issue with wireless I really like the Xubuntu user interface. It is clean, simple and elegant. I have been using Linux for years and I like this distro immediately.

I just purchased a Dell Vostro 200 desktop. Since I have been used to have dual monitor (laptop + extended monitor) I decided to have the same setting. But it didn’t turn out to be easy. Since I had to spend a lot of hours learning and researching, I want to write this post so you can save some time if you are in the same boat.

First of all, getting dual monitors for your desktop is easy in most cases. There are several ways that you can do it, and the most common ways are:

1. Get ONE video card with multiple ports (VGA, DVI). For those of you who don’t know the term, VGA is blue and DVI is white. You can connect an analog monitor to a VGA port, and digital monitor to DVI port. An adaptor is needed if you decide to connect an analog monitor to DVI.

2. Install 2 video cards.

However, in some cases (like mine), it is harder. Since I ordered a slim tower desktop, it is not easy to find a video card to fit my computer case. As a matter of fact, the video cards sold in the computer stores are not likely to fit. I tried both MicroCenter and Fry’s store and they don’t even carry this type of video cards. And I had a hard time to even explain my case.

Let’s ask my friend Google. One thing about searching these days is that you really need to know your keyword to search Google. For example, I tried searching “small video card”, or “video card for slim tower case”, and only got endless pages or generic online store results. After some research, I finally found their official name: “half height video card”.

However the search on “half height video card” didn’t yield much result. Apparently they are just rare.

Because the size restriction (half height, go figure), a half height card can only have one DVI/VGA port, and one s-video port at most, so the previous option No. 1 is out of the picture. But there is an alternative: use a video card that has DMS-59 port. With the DMS-59 port, you can use a Y-cable to split the signal with dual monitors. Here is an example of the Y-cable: http://www.amazon.com/Molex-DVI-cable-DMS-59-DVI-I/dp/B000BFGXOQ

Besides the option above you can always add a new video card to the system. But just make sure it is”half height” and “low profile” so it can fit the case.

One additional point is that choosing the correct card for your PCI slot. There may be different kind of PCI slots on your motherboard, for example mine has 2 regular PCIs, 1 PCI x1, 1 PIC x16. So getting the right card to fit the right unoccupied slot is another thing to watch for.

A couple of pictures.

I really should get an Apple some day.

Sometimes after making some (incorrect) network related changes I found myself locked out of the Linux server that I am trying to configure. A couple of examples are SSH configuration and iptables rules changes.

If the server is remote, say hosted by someone I’ll have to contact the support to have the change reversed and it could be time consuming and quite embarassing as well.

One way to undo the mistake is to schedule a change reversion just before the change and let it run a few minutes after. In this way, whatever change has been made will be changed back if the change was not successful. Of course if the change is good, I can cancel the job.

For the ssh config change, I can schedule this job before I restarted ssh service:

at now + 3 minutes <<< ‘cp /etc/ssh/sshd_config.backup /etc/ssh/sshd_config; service sshd restart’

A copy of sshd_config needs to exist of course.

For the iptables change, a similar job can be scheduled as below:

at now + 3 minutes <<< ‘iptables -F’

If you don’t like keep your system wide open you can always run this:

at now + 3 minutes <<< ‘/path/to/iptables_firewall_orig.sh’

Assuming the iptables_firewall_orig.sh exists somewhere (which is highly recommended).

For a forum builder, choosing the right software is a big step. I have used both vBulletin and Smiple Machine Board (SMF) to build forums, and some thoughts that I have maybe helpful to you.

Cost:

SMF is free. vBulletin costs $160. Both require to display their company name and licensing information but for $85, you can remove them from vBulletin.

Functionality:

Both apps offer great range of functionalities. A comparison head to head doesn’t really yield a lot of differences. But one major lacking from SMF is the missing of the WYSIWYG Editor. ForumMatrix provides a great comparison tool to illustrate this.

Modification and administration:

The biggest differences between the two are in this area. vBulletin’s admin interface is well-designed and much more complicated than SMF (also well designed), and as a developer, I feel SMF is more suitable for someone who has a strong coding background; versus vBulletin doesn’t require this. On this note, SMF gives more playground for flexibility and creativity if you are willing to dive into the code and work on it.

Software structure:

I like SMF for its clear design and separation of the core code and presentation. The core is stored in the Sources directory and presentation is stored in the Theme directories. In this way, Upgrade or changing theme will not affect each other. vbulletin saves all the theme changes in the database, and one can revert to the previous look if the new one doesn’t work out. Personally I don’t care much about this feature.

Performance:

SMF is known for its simple structure and faster load time. I don’t have any data to back this up. Hopefully as my forums grow I can have more info to add to this section.

This has puzzled me quite some time. Beginning a couple of months ago, I have been seeing some strange and broken symbolic links in my Apache home directory, which is /usr/local/apache. You can see them here:

lrwxrwxrwx  1 root root  25 Mar 29 23:04 build;460c7e42 -> ../../usr/lib/httpd/build
lrwxrwxrwx  1 root root  19 Mar 29 23:04 logs;460c7e42 -> ../../var/log/httpd

They are generated daily and each one has a different name. You can see how they are going to pollute the directory.

Cpanel is highly suspected for this. And I talked to the serve tech and the best answer I got is that the files might have been created by Cpanel or some other system procedures.

Finally I accidentally found a solution. I created a symbolic link to /var/log/httpd and named it “logs” and in this way the logs files and httpd.pid file will be written to /var/log/httpd directory instead of Apache home directory. Of course the original logs will need to be removed and renamed and Apache needs to be restarted. Guess what, the strange links stopped showing up.

This made me start suspecting logrotate since it apparently has something to do with the log files. But anyway, I am still have my fingers crossed and hoping this will fix it.

Let me know if you have the similar situation or solutions.

Backup is always an important chore for a webmaster. Recently I started to use rsync to backup important files between my servers. I don’t backup the code base since they are already in CVS. I mainly backup the database and some other user generated content, such as uploaded images, etc.

This is a great tutorial to use rsync to backup files and diretory trees.

The database content was first dumped into a file using msqldump utility by cron job, and later on another cron job will call rsync to sync with another server, which can be accessed through SSH.

So far this has been working great. The next step will be setting up script to periodically rsync the configuration files such as httpd.conf, my.cnf … so changes won’t get lost in case the server failure or migration.

I found this quote quite interesting:

In preparing for battle I have always found that plans are useless, but planning is indispensable.
– Dwight D. Eisenhower

As a webmaster, we also need a good planning skill. A good start with sufficient planning and avoid the extra works and continuous planning while running the site can help keeping the site running smoothly.

Here are some things I would do when starting a project:

1. Find the right hosting plan and host service.

The right hosting from beginning can pay off in the long term. I used some crappy hosting service off the start and I had to spent a lot of energy to worry about the downtime. I eventually dumped the service but I can’t get my lost time and productivity back.

If you want to run a quality site with a large code base the right hosting makes a lot of difference.

2. Design the site, page and more importantly URL

Most people focus on the graphic design at the beginning to make a site look sexy. Sexy site sells. But at least equally important, are the page navigation, content design and URLs.

A good page navigation organize the content hierarchically; the good use of on page html markups (h1, h2, etc…) make the page more readable to the search engines. A good markup structure is a good way of information organization.

A new site, for example a directory site with a bunch of empty categories, can possibly have a lot of empty pages that look similar in the Search Engine’s eyes. In Google’s case, they will be excluded from the main index and dumped into the supplemental list, which is a place a page should try to avoid getting into. A webmaster should plan on the growth and gradually add the pages naturally.

The URL is also an important piece. Ideally the URL for each page should be well considered from the beginning. Although a 301 redirect can always be done if the URL changes, it will be much better to start with good URLs and stick to them.

3. Plan on marketing the site

No matter how good your site is, you will need to get the work out somehow. There are a lot of ways to do this. And a good planning should use different marketing strategy progressively. This will make the advertising more effective, and your marketing $ goes longer way.

For a newly start site without much content, it is probably not a good idea to dive right into a big advertising plan off the start.

The early marketing steps may include post your site on different webmaster forums for reviews, submitting to free or low fee directories to get some low power back links.

As the site gradually improves it is the time to target links from some more authoritative sites.

Web site marketing is tedious and continuing process, the continuous effort is what it takes to make a name in a already crowded community.

Running a site:

1. Always watch your site.

A website is a 24/7 business and it should be monitored to make sure it stays up that way. There are some good free service for website monitoring and you can use a combination of them to monitor a site from different geographical location.

A frequent downtime is an early warning for an unreliable hosting service or a sign to investigate the code and optimize.

2. Planning for space and traffic.

Using Google Analytics, you can have a good idea about the traffic of your site. A webmaster should plan ahead for next hosting upgrade instead of waiting till the site can’t handle the level of the traffic. The increment of the traffic is always a good thing for a webmaster, but can a site handle it? The “Digg” or “Reddit” effect can bring a burst of fame to a site, but they can also take a site down.

Changing host is big deal but if planned early the execution can be smoother. For example, setup a staging environment early to regression test the site will eliminate most of the “surprises” from the actual migration.

3. Watch the system logs and attacks.

If you are using VPS or dedicated server, you have a great responsibility to keep your box secure for the sake of your site(s) and the whole web community. A webmaster should continue checking out the logs for the potential break-in attempt and misuse of the email system.

I am amazed by this video. It is simply fascinating. Just a thought, some day in the future the touch screen control will replace the touch pad and mouse and we’ll be coding for the “touching” events for websites.

Recently I had the chance to dip into some popular Javascript framework and tried out Dojo and Prototype. Even I only scratched the surface here is a little comparison from the experiment.

What are common between them:

Both Dojo and Prototype are Javascript framework or libraries. They simplifies the work of the developers by hiding the nasty stuff such as browser compatibility checking and providing simpler API calls to deal with the DHTML components.

Both of them support Ajax and make it easier to use.

Even Ajax is their selling point, both of them provide much greater capacity for writing other Javascript functionalities and not re-inventing the wheel.

Both of them encourage the clean programming sytax in a way.

Their differences:

Since Dojo is a toolkit, it contains a lot more libraries such as animation or flash. On the other hand, Prototype is closer to the ground and often used as the actual codebase to develop a library or toolkit. For example, Scrip.aculo.us is developed on the base of Prototype.

Based on the last point, Dojo has a much bigger package versus Prototype is essentially just one Javascript file.

Dojo appears to have better documentation but the Prototype community is working on creating better documentation.

My take on them:

Using a framework can make some work simpler but they may also cause other issues if you are not committed using them. That being said, both of the frameworks are nicely developed and useful. I guess the choosing of one over the other is based on where and how they will be used.

If I have a website that use some Javascript in a limited matter, I’ll definitely consider using Prototype. The main advantage of it is its small size and flexibility.

If I were to develop something that heavily relies on Javascript and most Ajax driven, Dojo may be a good candidate. With its widely range of widgets out of the box, it is possible to throw some fancy stuff together in a relatively short time.