A straw poll for people who look after websites...

[lostcarpark]

Lots of poeple look after different types of websites, whether it's a personal site hosted on something like MySpace, right up to a a corporate e-commerce site clustered across hundreds of servers.

If you find, or have ever found yourself in such a position, I'd like to know what's the one thing about the experience that ticks/ticked you off the most. Could you post a tiny bit about the site in question (no need to identify it unless you want to), your role, how the site is/was maintained, and either the thing that you found really annoying, or something you thought that would have made your life a lot easier.

This sort of ties into something I'm working on. I'm not promising I'll be able to make your life easier, but I'm not saying I won't...

Comments

[dougs.livejournal.com]

I always prefer to be the person who looks after the infrastructure, rather than the content, and I've not often had irritations in that role.

However, when I've been looking after content (which I do sometimes), the most annonying thing has always been people who have wanted some content changed, but have not been able to pin down what they want it changed to.


The things I've found annoying as a user are another matter.

[lostcarpark.livejournal.com]

I don't think I specifically said content, though I may have implied it. Feel free to include what bugs you about infrastructure too.

[hawkida.livejournal.com]

Comments like "It would be good if we could put something on the website about that" that are left hanging that vaguely.

No response on requests for wtf needs doing, and then suddenly you get a pile of work and it's urgent and the requester fails to allow for you having a real job and a social life that need to be juggled around volunteer work.

Insistance upon horrid things like blinking text and animated gifs and you're just a code monkey with no say in what is accepted.

These are past experiences. The current sites I maintain voluntarily don't really cause me any issues.

[lostcarpark.livejournal.com]

I can see where you're coming from. Often the technology is easy, but the people are the problem.

I notice you say "voluntarily". I'm interested in professional experiences too.

[hawkida.livejournal.com]

Oh professionally the worst are the clients who want it to look exactly the same on every machine and browser and just can't grasp that web pages aren't even *meant* to do that.

In my last job I'd spend ages getting stuff as the spec required and then they'd go, "Oh, we just want a minor tweak or two" and the tweaks although seemingly small took just as long to code. Then they'd "tweak" it back, or partially back and ask for incremental changes. Basically, the spec was a starting point and it was insane.

Worse was when we were bought by a global firm and given instructions. We'd go "But if we do it like this then x, y and z will happen" and they'd go, "Well, it has to match the spec because that's what the testers will test against. We can log the problems as bugs later." So you'd do insane stuff just to match an insane spec. This varied from calculation problems right down to spelling mistakes. You had to put in the bad spelling in order that it could be passed as identical to what was requested and *then* it could be fixed in the next release.

Then there was the point where I got put on bug fixing and I'd read the bug and say "Well, I don't know what page they were on or what they were trying to do and despite emailing and ringing them they won't talk to me" and my boss would say "Work around that". And then after I failed to work around knowing sod all about the problem *he* would contact the person who raised the bug and because he was senior they would give him all the info and he'd go "See? All you had to do was ask".

I don't think these are really restricted to the web, though.

[akicif.livejournal.com]

Ooh.... This comment box is too narrow. I'm going to have to have a serious think, and come back to this.

But I've just spent two days in a meeting (well, we did have three very nice meals out), so I'm off to the pub for now?

[akicif.livejournal.com]

THinking about this, it's maybe time for a friends-locked post going through my CV, plus the places I don't put on it, listing what wasn't so good.

Only thing is, if I include a link to my present day job, it'll show up in the referrer logs (they know about my moribund blog, but not the LJ)

[lostcarpark.livejournal.com]

Well, you could just give the name/site name, but not make it a clickable link, so people have to type it into the address bar, breaking the referrer link.

[hawkida.livejournal.com]

That doesn't work. The referrer is often just picked up as being the last page the user was on. If I'm on your site and type in some random address the new site will know where I was last. You'd have to go to a different site in between, or paste it into a different browser window or something.

[lostcarpark.livejournal.com]

Oops. I never realised that. I must look at my referrer logs more closely.

[spodlife]

My parents have a website for their business. The hardest thing is for them to understand the overall management of HTML documents, so now I have the whole thing stored in a Subversion repository. Before updates hit the website I massage the content to conform.

I want a tool or method which makes layout and common content (ie. menu bar) easier.

[lostcarpark.livejournal.com]

One of the problems in my experience is that the different elements of a website all tend to get mixed up together. So if you're a web designer, you also get caught up in programming to make things like menus work, and content management.

It's probably a little days of printing, where a typesetter needed to know all the details of how a printing machine worked so they could lay out type that wouldn't jam the machine, or leave black blotches on the paper. Nowadays the exact method of reproduction is of no consequence to the graphic designer, and the tools used abstract all of the details so he or she can get on with just designing.

We need to get to a similar state with web design. A web designer shouldn't need to know the intricacies of HTML or CSS, or how to get around the peculiarities of different browsers, they should be able just create a design and it should work. And they shouldn't have to worry about programming tasks like making menus or feedback forms or online shopping services work. It should just be a case of drag it on and the site will know how to make it work.

In many case someone entirely different will be maintaining the content, and they shouldn't have to worry about that stuff either, or wonder will their content render correctly in the designer's peculiar template. They should be able to just edit the content (or upload their hundreds of Word documents), and they should work.

And, then there's the whole technical end of servers and protocols and DNS mappings that end users too often get troubled with.

I think we've got a long way to go in this department. I've looked at existing CMS solutions, and I'm not convinced that any of them have got it right yet.

[spodlife]

I like LJ as a content management system, but that's because my content is a diary.

I have heard of people using wikis and then trying to hide the editability to normal viewers.

[lostcarpark.livejournal.com]

LJ is not at all bad for the types of content it's suitable for. And it does effectively separate the content from the design. It's not quite so good for separating the design from the underlying browser architecture.