Category Archives: Web Applications

Web Applications

What is FriendFeed?

I was just about to blog about FriendFeed as being the perfect aggregator for using specialised web services when I read the announcement that Facebook had bought the company (a more sincere form of flattery perhaps, than their original imitation) and it kind of put the kibosh on the post.

Now however, a while after the purchase FriendFeed still fundamentally has all the good stuff I wanted to write about and I’m still using it to aggregate more and more feeds so I felt it was worth picking up the topic again.

FriendFeed is a social networking site that has a lot in common with a lot of similar sites. You get microblogging with unlimited length comments (rather like Jaiku). This is handy for posting links and discussing them. From this fairly basic start you have two threads of innovation. The first is around the event handling, Robert Scoble probably has the best example (’cause he has a lot of readers) on his blog. On the right-hand side you can see the FriendFeed plugin and if lucky you can see that it is actually a real-time discussion widget. As people interact with the posts it gets feed onto the page. It’s pretty cool but also a bit of gimmick in a similar way as Wave is.

For me what is more interesting is the idea of “Imaginary Friends”. If you are a user on FriendFeed you can define the aggregation stream you want to present to the world. This is very similar to the aggregation done by Whoisi, except that you control the content in FriendFeed and the world decides on what your stream is in Whoisi. This is a handy feature but often you want to have an aggregation for someone who isn’t on FriendFeed or where you actually want to construct a different aggregation than that the user has chosen to present (for example excluding personal material from the feed). These aggregations are created for Imaginary Friends that you create within your own user context. My most common usecase is that I want to blend someone’s Twitter and Blog feed into a single stream so I can view their updates in one scan.

You can then create grouped sets of streams where you see a single feed that is an aggregation of aggregations. So for example I have a group of Imaginary Friends that represent pop stars on Twitter. I don’t need real-time updates from Lily Allen but on the weekend I do like to have a browse through and see what is happening with the Twitterati.

The final feature that is interesting (but a little creepy) is that you can use the FriendFeed commenting system to comment on other people’s feeds and have discussions without ever notifying the original poster. This allows side-channel communication but I do sometimes feel that it is more like talking about someone behind their back. This is probably most useful on things like Tumblr where there is no built-in comment system or a post where the comments or forums have been locked down.

So that’s my description of FriendFeed and while I was quite cool on it at first I have found that it has replaced things like RSS readers for me and I am starting to look at holistic aggregation as a necessary feature for information organisation systems.

Standard
Web Applications

Heroku versus GAE & GAE/J

At the Geek Night last night there was a lot of questions about how Heroku compared with Google App Engine (and I guess implicit in that how it compares to GAE/J with the JRuby gem). Since it came up several times I thought it would be interesting to capture some of the answers.

Firstly there are the similarities, both deal with the application as the base unit of cloud scaling. Heroku offers several services that are similar to GAE, e.g. mailing, scheduled and background jobs. Some aspects like Memcache are in beta with Heroku but fundamentally both services seem to concur.

They are both free to start developing and then the pricing for the system varies a bit between the two, GAE might be cheaper but also has a more complex set of variables that go into the pricing, Heroku prices on: data size, web processes required and then options, so it’s easier to project your costs.

Differences then: well a big one is that Heroku uses a standard SQL database and also has excellent import/export facilities. GAE uses Google BigTable which is great for scaling and availability but data export and interactions are very limited. Heroku has an amazing deployment mechanism that involves simply pushing to a remote branch. GAE is slightly more involved but it is still relatively easy compared with deploying your own applications.

GAE provides access to a variety of Google services that are handy such as Google Account integration and image manipulation. GAE also has XMPP integration which is pretty unique if you have the requirement. However these do tie you to the platform and if you wanted to migrate you’d have to avoid using the Google webapp framework and the API. If flexibility is an issue you are probably going to prefer Heroku’s use of standard Ruby libraries and frameworks.

Heroku has few restrictions on what can and can’t be done in a web process, you also have quite powerful access to the user space your application runs in. GAE and particularly GAE/J has a lot of restrictions which can be quite painful particularly as there can be a discrepency between development and deployment environments.

However running a JVM in the cloud provides a lot of flexibility in terms of languages that can be used and deployed. Heroku is focussing on Ruby for now.

From my point of view, choose Heroku if you need: uncomplicated Rails/Rack application deployment, a conventional datastore that you want to have good access to, a simple pricing model. Choose GAE if you want to have out of the box access to Google’s services and a complete scaling solution.

Standard
Ruby, Web Applications

Haml, Sinatra and template variables

Haml advertises itself as offering a haiku for HTML; that might mean it is short, beautiful and expressive or it could mean it is enigmatic, unsatisfying and cryptic. After a period away I have forgotten how Ruby-centric Haml is. If you want to include parameters in a Haml template within Sinatra you need to bind a map to the parameter locals.

haml :index, :locals => {:hello => “World”}

The easiest way to then include the data within a tag seems to be to use Ruby string interpolation

%p= “Hello #{locals[:hello]}”

There may be more elegant ways to do this but it seems relatively idiomatic. It is just hard to adjust to after months working on very restrictive templating systems.

Standard
Python, Web Applications

Juno, a micro web framework for Python

I love microframeworks and I love Python to it therefore follows that I love the minimal web frameworks you can get for Python like CherryPy. However as soon as I heard about Juno, a clone of my favourite microframework Sinatra I had to give it a go.

For me the archetypal application for Sinatra is Git-Wiki and therefore to kick the tires on Juno I decided to try and do the same thing with Bazaar. Hence:  Bzr-Wiki on Juno!

You need Juno, Jinja2 and SQLAlchemy as well the source from Launchpad. Once you have it all then got to the repo directory and type bzr init. You should then be good to fire up wiki.py and muck around with it.

It is slightly ugly and lacks access to revision history, diffs and user tracking but apart from that it is a surprisingly functional wiki. It is also bi-directional in that you can add files to the Bazaar repo and they get reflected in the app.

So what was Juno like to work with? Well overall I thought this was the best Python microframework I have used so far. I really like the idea of decorating methods to avoid having to generate a mapping table. The syntax is terse and comprehensible, the conventions around the framework made sense to me.

By comparison with its progenitor I think I really missed the autoloading/dynamic evaluation that allows you to change code in Sinatra and have it immediately take effect. The function of the Request decorator was initially quite obscure (it binds all HTTP verbs to that method, if you want to map GET to another method you must specify all verbs independently) and I am still not sure it is right. I think the most specific decorator should take precedence. Other than that I think the framework ports a lot of the concepts from Sinatra in a sympathetic way.

The dependency on SQL Alchemy is also really clunky. If you specify that you are not using a database (as is the case here) then it is annoying to have to download a dependency and makes installation on Windows a pain I didn’t even want to try and tackle.

Juno is really promising though and I look forward to it developing. I think it would be a real delight to use it in an environment like Google App Engine.

Standard
Programming, Ruby, Web Applications

Why does Heroku matter?

Heroku is an amazing cloud computing service that I think is incredibly important but which really isn’t getting the visibility it should be at the moment.

Why does Heroku matter?

Firstly it gets deployment right. What is deployment in Heroku? It is pushing a version of your app to the cloud. That’s it. The whole transformation into a deployable application slug is handled by the Heroku, pushing it out to your dynamos is handled by Heroku. Everything is handled by Heroku except writing your application.

Secondly, it is the first cloud/scalable computing targeted directly at Ruby that has the right pricing structure to encourage experimentation and innovation. It has the low barrier to entry that Google App Engine has but requires no special code and imposes no special restrictions.

Thirdly it does the right thing with your data. It respects your right to access, mine and manipulate your data and provides a mechanism for doing this easily.

For me Heroku has set a new benchmark for Cloud Computing and build and deployment in the enterprise as well.

Standard
Ruby, Scripting, Web Applications

Semi-static CMSs

You know you have a good idea when there are half a dozen implementations of it. All essentially the same but filled with subtle nuance. The latest idea I have stumbled across is the “semi-static” website generator. The idea is so painfully obvious I cannot believe that it has taken us until 2009 to do it properly.

What is a semi-static website? Well it’s actually something like this blog. I spend ages writing and editing the articles but once I publish something I don’t generally change it unless it has factual errors. I need something that allows me to make changes but my content easily but my content doesn’t really benefit in anyway from being stored in a database or being generated dynamically on the fly because it is going to spend most of its life not being changed.

If you look at a lot of CMS systems the same is true. People want to be able to edit and reorganise material easily but they are rarely changing it more than once a week. The majority of the time the pages are static.

If you look at a number of CMS or Wiki systems they are actually quite complicated. Before you can get going and generate your content (which remember is the core of what you want to do) you might have to setup a database and get the application deployed to a webserver.

What the semi-static website generator does is provide a set of scripts that allow you to edit and create content and then “compile” your site. The result is set of static HTML, Javascript and CSS files that can be deployed to your static webserver (which is generally pretty easy to setup, most servers will come with Apache all ready to go).

When you combine it with a bit of DVCS magic deployment is even easier. You just checkin your new pages and then update the webserver copy of the content.

So if you want to get into this kind of content management what options do you have? Well you have a lot of choice indeed. Almost all of them are Ruby projects and they have a certain similarity to the Rails concept of code generation. You will be issuing commands like “create page”, “build site” and so on. You operate at one remove from it.

There are many, many, many choices but for myself I chose three to look at Nanoc, Webby and Jekyll. Nanoc I came across via Proggit and was struck by the simplicity and clarity of the documentation.

Installation of all three is via gems but there are some differences here. Nanoc is a very simple Gem and works with HTML and Erb by default, if you want to use a markup language or different templating language you will have to install the gem yourself. This means it also installs easily under JRuby too.

Webby by comparison is a very noisy install, I think it may have its dependencies slightly wrong as it installs a version of rake, flexmock and rspec which are surely development dependencies rather than anything I as a user would need to use for the framework. It also uses hpricot which means you can’t install it under JRuby.

Jekyll has a similarly demanding set of requirements but will install under JRuby. Jekyll did feel a lot like a personal solution as I was installing it. Textile is great but if I am going to be using Markdown why do I need to install RedCloth?

Overall I liked the nanoc install experience most.

Moving on to using them. Nanoc and Webby are very similar in that they use generators to create new elements in the site. Jekyll seems to use convention instead. At this point it is worth mentioning that there are some big differences between the three products in their intentions. Nanoc is a general website creator but it does have a certain amount of opinion as to how to do that. For example it generates very nice clean website urls by putting each page in a directory by itself and then creating the page as an index.html page within that directory. Make sure you have your webserver configured appropriately.

Webby is pretty flexible but comes with some built-in templates that will be build you a basic page, a tumblog style site, a normal blog and a presentation. That was a pretty neat feature and one the other frameworks didn’t have.

However the negative side to this is that the templating is based on Blueprint; which is both excellent as Blueprint is a great way of designing pages and a pain as Blueprint is actually pretty involved in terms of understanding the grid concept and what the various span-x css classes translate to. I felt that Webby did a great job of offering cross-browser styling that looked good but that in doing so it didn’t make the simple stuff simple. With my trial site my sidebar came out on the bottom of the page and I didn’t really know how to get on the right-hand side of the page. In nanoc I could just float: right it, in Webby I felt I had to understand the grid that had been created for me. I’m not sure it really allowed you to use any CSS knowledge someone might already have but the benefit is you know your site is going to look good across browsers and in print and so on. For more professional site development this might actually sway you to select Webby.

Jekyll is primarily a tool for publishing a blog style site so it has a lot of built-in support for special information relating to posts but in terms of layouts it uses general HTML templating like Nanoc does. It allows  for the creation of little chunks of reusable content across pages via includes but lacks some of the flexibility of Nanoc’s tag helpers. While focussing on blog publishing Jekyll actually encompasses a powerful set of processing tools for content that effectively allows you to use it as CMS. Jekyll is used to process GitPages on GitHub, a real demonstration of its power beyond just blogging.

For what I wanted to do I initially felt that Nanoc was the best choice, it’s flexible, powerful and generates nice sites. Webby is powerful but didn’t expose that power to me in a way that made me feel in-control. However when I started to get to grips with Jekyll I was really amazed at how powerful the use of a few conventions were. Jekyll processes your site’s pages effectively in-place. It can choose to process your files if they match a naming convention (such as having a content extension like textile or markdown) and it uses an embedded YAML section in the file to do things like provide the title of the page or select the layout that the content should be displayed within. While Nanoc produces fancy URLs Jekyll made it easy to build more conventional websites consisting of clusters of pages.

Jekyll uses a small number of special directories (indicated by a leading underscore) to hold things like layouts, cross-page snippets and posts (if you are using the dated posting functionality). However for a simple site the number of directories and configuration files was significantly less than Nanoc. In fact you could get by with just a master page layout, which would require just one special directory and page within it. Everything that should have been simple was with Jekyll and the result was very close to a conventional website you might have coded by hand but extremely fast to create. It also uses just one command to press the pages and fire up a local server to view the pages. Both Webby and Nanoc seemed to make pressing and previewing slightly harder than it needed to be and at least once I forget to execute the Nanoc commands in the right order.

So, conclusions… Going forward I will be using both Nanoc and Jekyll. Webby I would be willing to consider but I don’t really have a task for it that would reward the additional effort I have to invest in its capabilities. If I just want to bang out pages for a website but minimise my effort and make it easy to abstract common content then I will definitely be using Jekyll. It is a joy to use. For sites that are maybe more based around articles or grouped content then I think I will be using Nanoc. The default URL construction is nice for hierarchical data (as long as you can stand the many directories) and there’s a sense that the configuration files make it easy to make changes that affect the site as a whole. However it does use a lot of scaffolding to achieve the effect and there are a lot of exposed options that could have been reduced by adopting some conventions Jekyll style.

Standard
Java, Programming, Python, Scripting, Web Applications, Work

JMeter or The Grinder, so which one is better, like?

Or for the benefit of Google: Apache JMeter versus The Grinder. Fight!

A while ago I got paid to put these two tools head to head and I think it’s been long enough that the people who put up the money will have got the benefit now.

I had used The Grinder in a previous job where it had done excellent service finding out what level of peak load our site could handle. It was convenient in that you script it in Jython because Jython was our chosen scripting language.

JMeter on the other hand was a whole new thing to me. It is a kind of graphical programming interface where you drop controls into a tree structure and the framework generates threads that run through the tree generating interactions with the target application via things like HttpClient.

The scripting versus component structure is a pretty major difference between the two and is probably one of the key things that is going to help you choose between two products that are both, to be honest, pretty mature, capable pieces of software.

If you don’t feel comfortable with Python or Jython then Grinder is likely to be out. JMeter is much more friendly to non-programming testers. However this decision isn’t as clear as you might think, JMeter actually includes Javascript functions, access to BeanShell and its own expression language. As we created more complex traffic simulations we found ourselves having extremely complex expressions that often mixed a scripting language with an expression evaluation. In some ways The Grinder is more honest in telling you up front that you are going to have to program some of this stuff yourself.

Another big differentiator is in reporting, if you need to generate reports then JMeter is a much better choice as it has components that collect and collate information and you can dump data to XML for XSLT transformation into pretty much any output format you want. It is also possible to run JMeter headless once you have the test scripts so it is possible to encorporate it into a continuous integration process. Debugging slightly flaky applications is much easier with JMeter because you can easily capture a lot of result information and then just delete the data gathering component once the test flow is working.

A lot of the functionality of both products overlaps: they can both be used a browser proxy and be used to capture scripts as a user “drives” around the site. These captures are likely to form the basis of your test plans unless you have a very clear URL scheme with little session state.

Both use thread pools to generate load and both tend to get blocked on your application before they max out their local machine (although both are capable of using 90% of memory and CPU). JMeter has a few nice options around randomly ramping up the thread activation (recreating a spike in usage more closely). The Grinder tends to give more bang for the buck as its runtime is very simple and minimal.

Both are also capable of using distributed agents and collating the data from these agents back to the coordinator.

In many ways there isn’t a “bad” choice to be made between them. So what might sway your choice one way or another? Well JMeter is an Apache project and that might make a difference for some people as you have all those good things like project governance and a good chance the project is going to continue to be developed and enhanced. JMeter was also the only project to have a test suite last time I checked out both code bases.

The Grinder does have one unbeatable quality in my opinion and that’s flexibility. When you look at the features listed on the websites you might think that JMeter does all this cool stuff with HTTP, SOAP and POP3 and the like and that The Grinder is comparatively light in the feature set.

The opposite is actually true, as The Grinder website points out the The Grinder can test anything with a Java API. There really is no limit to what you can make it do or how you can execute your test plan. In fact most of the things I’ve complained about with The Grinder are fixable from within the test scripts  (what I am really complaining about is that some things like capturing HTML output per run is something that should be available as part of the standard package). On the other hand I felt that creating a new JMeter component was actually quite tricky as you have to deal with both the GUI aspects and the actual functionality you are trying to create in your test component.

If you really want to have total control of your concurrent volume testing then The Grinder is probably the best product for you.

Perhaps the last topic for consideration is The Grinder’s use of Jython. Python is a great language and you get a lot of power for some very concise code. Just as some people are going to find it off-putting others are actually going to find it very compelling.

Okay so yet another weak “horses for courses” post on the InterWeb. Perhaps the easiest summary to end on is that your test teams will probably take a shine to JMeter while your developer teams who are also building scale tests are probably going to like The Grinder; unless they dislike Python or learning a new language.

Standard
Python, Web Applications

Django in 24 hours

Last Saturday afternoon I decided to learn Django. It was 2pm on the first day of SiCamp 2008 in London and being the only developer in the room at that point I decided that I should do whatever I felt would be the best option to get an application running by 2pm the next day.

Previously I have done some Google App Engine and the experience convinced me to give Django a go after I found myself, by intuition, creating a GAE project structure of handlers.py (views), models.py and a directory called templates that contained templates. I was then disappointed to find the whole world had got there before me.

So, Django in 24 hours, baptism of fire. What do I think now looking back on the experience?

Everyone has told me that the Django documentation is good and I think I have to concur. Not everything is so clear that when you’re speed reading in one window and typing in another it works first time but importantly nothing in the documentation is actually wrong. When stuff is not working a second, careful look at the documentation got me back on track.

Importantly Django’s core model of web development is sound and intuitive. My editor had around ten files open for the project and the flow of adding something to the application did naturally flow from url to handler to view to model. Maybe the only quibble I have is that the views.py file is deceptively named in MVC terms.

The core of the framework is amazingly concise, I spent the majority of my time thinking about the problem not about the framework API. Binding a URL to function made sense, having to specify a template instead of having one inferred from the method name was maybe my one criticism in the method handler but on the plus side it does allow for flexibility in handling requests. Handing off from one request handler to another was very easy.

Django templates are both amazing and annoying. The syntax and principles are amazing, it was easy to play around with the pages and the template inheritance was really powerful for avoiding duplication. However when I transferred the application from self-serving to mod_python the template generation was very wobbly when compiling changes from the file system. Of course this could also have been mod_python but it was the latest 3 series stable source compiled for the machine. I’ve used Jinja2 previously when generating HTML in Python and might be tempted to stick with it in future.

Django models are great, I hate ORM but I really liked syntax for defining persistence properties and I liked the way that you don’t have the fact that you are really dealing with SQL hidden away from you. It genuinely seemed a more convenient way of expressing the data model rather than an OO wallpaper over relational data storage. I didn’t feel the need to add domain logic to the models but I felt like it wasn’t really polluting the model to do that either.

One thing that didn’t work at all was changing the relations between models; it took me two or three attempts to finally model the relationships between the data concepts. Each time I changed a Foreign Key or Many to Many relationship I ended up deleting the database (SQLite3) as I couldn’t figure how to migrate from the old schema to the new.

One reason for choosing Django was the idea that I wouldn’t have to write the backend code as the admin stuff would be right there for me. It took me a while to get the 1.0 admin to fire up but once it was running it did perform as advertised. One of the attractive things about the application was that the data model followed the conceptual language of the solution in a really powerful way. You could use the admin interface to have a Devotee perform Devotion to an Action. My geek excitement peaked anyway, YMMV.

So them’s the highlights of the experience. Overall Django delivered me a rapid web development process in an intuitive, powerful way and lived up to nearly all of the claims made on the tin. Deploying to Apache/mod_python was painful but most of the pain surrounded the infrastructure of my box (multiple versions of Python, Apache config files) and my lack of mad Apache admin skillz.

I would happily tackle another project in it again.

Perhaps of interest is how the Django development experience matches up against Rails or GAE  which would have been the other obvious choices. GAE would have been very similar but the deployment would have been better and I wouldn’t have had any automatic admin. In retrospect it may have been a better choice for a hack party type event. It certainly would have been my choice for personal projects for easy of deployment but now I have one Django app running perhaps that isn’t as relevant any more. Certainly the thing that has kept me from GAE before, the pain of data migration, doesn’t seem that better in Django (except that you control the datastore and its contents).

Compared to Rails?

  • Admin is much more awesome than scaffolding.
  • Django ORM is much less complex than Active Record, all the data required to create, deploy and use the object is in one place. Django doesn’t have Migrations but has its own brand of database versioning pain.
  • RSpec is awesome (despite its monkey patching of Object) so you aren’t going to beat Rails for easy testing.
  • Django templates are more powerful and easier to use than Erb but you have a lot of Ruby templating options so it’s hard to make a complete comparision. They probably both similar in the sense that you can find a template library that suits your preferences. Django is the purely solution out of the box.
  • Routing and Controllers are much less involved in Django than Rails.
  • Django is less opinionated about how you structure your application directories, which I like.
  • Django doesn’t bake-in AJAX components but is “batteries included”, Rails probably generates better Web2.0 style apps for less effort.
  • Finally Django uses only a few code generators because its basic structure is far less involved. It also generates far less “stuff” for each MVC element which I quite like as I don’t tend to use everything Rails generates.

Okay detailed analysis over, what’s the high-level view? Django and Rails are similar experiences but I think the major differences between them are almost what you could say about Python and Ruby. In Django you are going to get simplicity, clarity and a real choice of how you plug your infrastructure components together. In Rails you are going to get magic up front which is cool but also magic at the back end, which is not cool. Ultimately I think the answer is how opinionated you like your software. Well punk? How opinionated do you like it?

Standard
Java, Web Applications, Work

Better, Faster, Weaker

I went to the Developers and Startups Meetup last week (look it up on Meetup if you’re interested it was a good event). One thing that struck me was that all the entrepreneurs were mostly interested in PHP and Ruby on Rails (there was a significant rump of ASP and .Net which I can only put down to C#’s language hotness at the moment). Java in particular was poorly regarded.

I was intrigued by what was causing the negative vibe as while I can understand that Ruby on Rails is the new tech hotness I couldn’t really see the appeal of PHP over something like Java that has a decent stack that is practically free. Talking to a few people I got the sense that PHP was seen as a “getting things done” language. Something that got you to a tangible real product very quickly. Rails seemed to have much the same quality but I think a lot of people felt that Rails skills were more expensive and in a lot of cases they wanted a thin web layer over an existing service backend that already existed.

Of course both PHP, Rails and the Java Web Stack all make various compromises to do what they do. Where Rails is opinionated, Java probably isn’t opinionated enough. If you put all your code into page (Model 1 stylee) then I’m pretty certain that writing JSPs in J2EE 5 is probably as quick to put something together.

Of course one thing that hugely hampers Java as a productive web stack is the compile and deploy loop. There is none of the immediacy of something like RoR that allows you to make changes very quickly in response to feedback. The tradeoff is of course that the code needs to be interpreted and you have to add some kind of caching to avoid the performance hit of always having to evaluate code.

But given that there are a group of people who need to have code quickly so they can sell something to their customers or investors and that they are willing to make all kinds of compromises about scale and maintainance costs is the Java platform really out of the question? I ask this because I think I have been involved in quite responsive Java web teams where feature changes have been matters of days not weeks or months.

I think that perhaps the issue is partially just mental. Enterprise software development tends to take place in a risk-adverse environment where certainty or process is valued over rapid delivery. “Best practice” in this environment tends to sacrifice timeliness. In the smaller business space this might well be a practice turned into dogma.

That said though some Java Frameworks introduce major barriers to just getting stuff done. Any framework that forces you to produce XML based interaction flow is probably distracting you from generating functionality that people want and will pay for. Routing isn’t a trivial part of an application but that’s all the more reason to box clever with it. Any framework that doesn’t make it simple to inject services into classes is making life too hard. Ditto any framework that forces you to work with a particular Expression Language or Templating product.

I am wondering if what I am looking for is actually the weakest Java framework. Rather than having functionality for all circumstances (and having to configure and setup it all up) I want a very basic REST framework which I can then add templating and persistence according to my need. It still wouldn’t have the nice scaffolding of RoR but it would mean that each element of the web app would be absolutely vital to its function. Weakness would generate the power of vitality and reduce the obfuscation of bloat.

Standard
Web Applications

Important! !important is a danger sign

Until recently I had never seen the CSS keyword !important used in a production site. However just recently I have seen it in use and also had to use it myself to fix a few cascade issues.

CSS selectors work by assigning a “magic number” that indicates how specific the selector is in relation to the other selectors. Important works by boosting that number by a magnitude or ignoring all other selectors entirely. You can read the exact rules in the specification.

Important is really powerful and as a result you never want to use it. It’s kind of like the CSS A-Bomb, if you ever have to use it, something has gone wrong. The biggest problem with !important is that it can “lock” a style element and make it hard to override it in other cascades. Inevitably this becomes a problem because there is pretty much nothing in CSS that can be regarded as universal in the appearance and rendering of a website.

This then leads to other stylesheets also using !important in their selectors to overcome the earlier !important. This limits their reuse as they now, in turn, export their overly powerful rules and therefore require yet more !important use and so on and so on until every selector has !important on it.

Stylesheets should try to have as weak as possible selectors (without going overboard and perhaps applying some styling information too liberally). This makes them more generally useful as often people only dislike a few elements in a style or an individual page only has a few components that do not gel well with the general style.

I think !important should never be used when creating CSS. There are perhaps two exceptions I can think of: firstly client styles, you know best, fill your boots; secondly, stylesheets that you know represent the real bottom of a cascade. For example an optional stylesheet that renders the site in monochrome can reasonably be expected to represent the final word in a cascade.

Standard