Should we let .html be used?

[edwardhorsford]

Should be let users have .html on the end of their route?

I recently tried to access a page, and mistakenly included .html on the end, as I copied the file name in to the browser. This feels like it could be a mistake other makes, and something we could take care of in our routing.

[joelanman]

I've not had this issue personally - anyone else?

[timpaul]

No, but I watched Ed do it ;-)

Are there any problems doing this could cause?

[joelanman]

well, any new code adds maintenance work - we need to know there's real need. hmm, we'd need to be careful that it still worked when people wrote their own routes or you could get weird behaviour.

[edwardhorsford]

I just watched @aduggin make this same mistake.

I reckon if .html is applied, we should redirect to the same url without .html. If that route exists, it will work, else it will 404 as it should.

[joelanman]

that sounds reasonable to me

[timpaul]

What happens if you have a directory called 'foo' at the same level as 'foo.html'? Or is that basically something you should NEVER DO?

[edwardhorsford]

@timpaul I'm not sure that's relevant? Right now if you go to localhost:3000/foo.html you won't get foo.html - it will error instead. This suggests removing the .html bit and searching that instead. Which will then fall back to whatever behavior we currently have.

[tsmorgan]

@edwardhorsford I'm with @timpaul. If you redirect localhost:3000/foo.html to localhost:3000/foo and leave the prototype with the behaviour we currently have, then you could have a situation where what you end up displaying is the template at /app/views/foo/index.html.

To really make this work you'd want to detect .html on the URL, remove it, but in those circumstances prevent the kit looking for /app/views/foo/index.htmlas a backup.

[joelanman]

That behaviour seems OK to me? The redirect will drop .html, so I don't think it'd be too hard to figure out what was happening?

[timpaul]

Yeah, I'm guessing you really should avoid having folders and files with
the same name at the same level anyway right?

On 8 February 2016 at 21:02, Joe Lanman [email protected] wrote:

That behaviour seems OK to me? The redirect will drop .html, so I don't
think it'd be too hard to figure out what was happening?

—
Reply to this email directly or view it on GitHub
#102 (comment)
.

[joelanman]

currently the file wins out - it gets checked first

[tsmorgan]

Yeah I see what you mean. You're not going to be left looking at localhost:3000/foo.html in the address bar and actually seeing the /app/views/foo/index.html template. I'm sure you're right it's unlikely to confused people in that respect.

[edwardhorsford]

I just saw @finiteattention make this same mistake, adding .htm and then .html to the end of her page.