Ideas for the new catonmat.net website(catonmat.net)
catonmat.net
Ideas for the new catonmat.net website
http://www.catonmat.net/blog/50-ideas-for-the-new-catonmat-website/
31 comments
Thanks for the feedback.
About the 301 redirects - it doesn't take much effort. People don't make mistakes too often, but when they do I wouldn't like to lose visitors that came from their link. I have currently fixed a few links that I have noticed via URL rewriting, but it's web server specific, and I want to brink it to application level to have all the site configuration centralized from admin menu. Managing this tiny feature won't take much of my time and I will still be able to produce quality articles.
About the job board. Oh, didn't think of that. You are right. I will set this idea a lower priority and just keep producing good content until I hit critical mass. :)
About the 301 redirects - it doesn't take much effort. People don't make mistakes too often, but when they do I wouldn't like to lose visitors that came from their link. I have currently fixed a few links that I have noticed via URL rewriting, but it's web server specific, and I want to brink it to application level to have all the site configuration centralized from admin menu. Managing this tiny feature won't take much of my time and I will still be able to produce quality articles.
About the job board. Oh, didn't think of that. You are right. I will set this idea a lower priority and just keep producing good content until I hit critical mass. :)
Personally speaking, my biggest complaint about your site is that the Lecture Notes you publish are for courses I have already taken online-- I only wish you had gotten there first. Damned space-time continuum!
Oh no... /o\ When did you take take these courses?
I did them both about 2 years ago. I didn't take any organized notes, I'm afraid-- I'd have been happy to pass them along if I had.
If you're going to continue following in my footsteps, I suppose the UC Berkeley Operating Systems course and Knuth's "Musings" series will be next up.
If you're going to continue following in my footsteps, I suppose the UC Berkeley Operating Systems course and Knuth's "Musings" series will be next up.
You're right. Next up is Knuth's "Musings". :)
The 404 page resulting from a request for www.catonmat/artikle could say "did you mean: www.catonmat/article".
This is an awesome idea. Wrote it down. :) But I will want to correct that URL transparently to the user if I know what it should be.
It doesn't happen too often. I could even automate some of the patterns. One of them is when you put your link in parantehsis like this (www.domain.com/path), some regular expressions include the trailing ) in the path and the resulting URL is 404. This can be fixed automatically by dropping trailing ')'s.
It doesn't happen too often. I could even automate some of the patterns. One of them is when you put your link in parantehsis like this (www.domain.com/path), some regular expressions include the trailing ) in the path and the resulting URL is 404. This can be fixed automatically by dropping trailing ')'s.
If you can get a complete list of valid URLs, you can try to fuzzy-match the URL against that. If one real URL scores much better than all others, you can redirect, otherwise provide a 404 with a choice of the top matches. I'm not sure automatically redirecting from incorrect URLs is within the HTTP spirit, however.
Good idea. I wrote this one down as well!
In addition to the above fuzzy-matching idea, you could also do something similar to Wikipedia, and for redirected pages, included a notice at the top like:
Redirected from catonmat.net/artikle
...to let the user know that they did not type in the most preferred URL, so hopefully they will correct their mistake in the future.
Redirected from catonmat.net/artikle
...to let the user know that they did not type in the most preferred URL, so hopefully they will correct their mistake in the future.
Also a great idea. Thanks!
I don't think no. 30 is a good idea, I find it very irritating when a site does this. If I want to open something in a new window/tab, I'll do it myself.
I'm not keen on it either. I think the behaviour of the browser should be that the user controls whether it goes in a new window or tab, i.e.: LMB = same tab, MMB = new tab.
Oh, and if the link is Javascript that attempts to enforce new or different tab, then thre browser should send a distributed denial of service attack to the website. Or better still, assassinate the person resonsible for it (though I suspect this is AI-complete).
Oh, and if the link is Javascript that attempts to enforce new or different tab, then thre browser should send a distributed denial of service attack to the website. Or better still, assassinate the person resonsible for it (though I suspect this is AI-complete).
Thanks. I will reconsider this idea.
I think you should focus on improving the design of your site. Make it more simple so your content is focused - remove the weird matching colors, cloud picture taking too much space on the top and other things that take the focus of your site's main attraction: great content.
I think you could get inspired by A List Apart ( http://www.alistapart.com/ ), they provide similar articles like you, but their design and typography is much better.
I think you could get inspired by A List Apart ( http://www.alistapart.com/ ), they provide similar articles like you, but their design and typography is much better.
Thanks for the suggestions, Amir. Web design is not my strong side, but I will try my best!
I really, really recommend working with a web/graphic designer in that case... I always used to go solo as I wanted to retain complete control of the finished product. What I do now though is I consult with a graphic designer. He may not make even 50% of the design, but his input always makes a huge difference in the visual quality of the finished product.
Thanks, I'll think about it.
Hmm...
Who changed the title from "50 ideas for the new catonmat.net website" to "Ideas for the new catonmat.net website"?
I liked the "50 ideas" more. As tests have shown, people like concrete number of things, like "10 Things You Should Buy For Christmas" or "5 Unix Utilities Everyone Has To Know"
Any way to change the title back to "50 ideas"? :)
Who changed the title from "50 ideas for the new catonmat.net website" to "Ideas for the new catonmat.net website"?
I liked the "50 ideas" more. As tests have shown, people like concrete number of things, like "10 Things You Should Buy For Christmas" or "5 Unix Utilities Everyone Has To Know"
Any way to change the title back to "50 ideas"? :)
There's a HN guideline that says:
If the original title begins with a number or number + gratuitous adjective, we'd appreciate it if you'd crop it. E.g. translate "10 Ways To Do X" to "How To Do X," and "14 Amazing Ys" to "Ys." Exception: when the number is meaningful, e.g. "The 5 Platonic Solids."
That's probably the reason why it was edited out.
If the original title begins with a number or number + gratuitous adjective, we'd appreciate it if you'd crop it. E.g. translate "10 Ways To Do X" to "How To Do X," and "14 Amazing Ys" to "Ys." Exception: when the number is meaningful, e.g. "The 5 Platonic Solids."
That's probably the reason why it was edited out.
Oh, thanks!
I had missed this HN guideline.
Sorry about that, will not repeat this mistake!
I had missed this HN guideline.
Sorry about that, will not repeat this mistake!
People like it because lists are easier to digest than ideas embedded in prose. It's also easier to write, so it's a symptom of sloppy articles.
I'm intrigued by the Knowledge Database idea, care to explain a little more about that pkrumins? I may build something similar for myself.
Sure.
I had a knowledge database back in 2002-2004. It was an application that I used every time I was learning something. It wasn't too advanced, I had to do everything manually, but now I am going to automate as much as I can.
Here is how it worked, every time I'd learn something from the web, I'd copy/paste the key information into the application from the source I was reading, add tags, wget the page (so that i had local copy of the database) and link to local copy. This was I was accumulating tons of information. Some days I'd review everything I had accumulated and add some comments with where it could be relevant. When I needed something, I'd just enter some phrase or keyword that I remembered, like 'linux security chmod' and I would locate the key facts about this topic that I had stored.
Now that I think of it, it can be argued that you can do the same with Google, but it would be a bit slower, you'd have to see which site was it and you'd sometimes make mistakes of locating the right one. But if it's in the database, there are just a few results and you can recall precisely which one is the right because you added the result. It's you personal knowledge database, not the knowledge database of the whole universe as Google is.
I had a knowledge database back in 2002-2004. It was an application that I used every time I was learning something. It wasn't too advanced, I had to do everything manually, but now I am going to automate as much as I can.
Here is how it worked, every time I'd learn something from the web, I'd copy/paste the key information into the application from the source I was reading, add tags, wget the page (so that i had local copy of the database) and link to local copy. This was I was accumulating tons of information. Some days I'd review everything I had accumulated and add some comments with where it could be relevant. When I needed something, I'd just enter some phrase or keyword that I remembered, like 'linux security chmod' and I would locate the key facts about this topic that I had stored.
Now that I think of it, it can be argued that you can do the same with Google, but it would be a bit slower, you'd have to see which site was it and you'd sometimes make mistakes of locating the right one. But if it's in the database, there are just a few results and you can recall precisely which one is the right because you added the result. It's you personal knowledge database, not the knowledge database of the whole universe as Google is.
If there was a web app that did this would you use it?
And the killer question: would it be worth a monthly fee if it really rocked and had desktop/iPhone apps?
And the killer question: would it be worth a monthly fee if it really rocked and had desktop/iPhone apps?
Not sure I would use it. I would need the knowledge database to be customized exactly to my needs. Perhaps other people would use it.
Here is a concrete example: Someone links to www.catonmat.net/artikle when they wanted to link to www.catonmat/article. I’d simply insert an entry to 301 redirect /artikle to /article and everyone’s happy.
Everyone's happy? I think not. What you've just done is taken on the the onus for fixing other people's mistakes. In the long run, that's not sustainable-- it's more work for you, and only encourages sloppiness on their part. It's not a road I'd like to go down, I'll tell you that.
47. Add A Job Board. As my site is getting more popular and popular among programmers, it may be a good idea to add a job board. Joel Spolsky made a million $ in a year with job boards. As the popularity of my site increases, I might make a few dollars out of it as well.
The value (and revenue) from a job board is not linear-- if your site gets 5% of the traffic that Joel gets, your job board is not going to have 5% of the value. In fact, until your site is extremely popular, a job board has almost no value whatsoever. It takes a critical mass.
Catonmat's a great site, btw-- I really dig your content. And that's why I'd hate to see you putting your energy into low-value, high-effort activities like the aforementioned, instead of concentrating on the core: producing good content.