Hacker News new | past | comments | ask | show | jobs | submit login
Convert Any Webpage to an Image (linkpeek.com)
73 points by foxhop on Dec 5, 2011 | hide | past | favorite | 71 comments



It's been 6 hours since this was posted and there are still NSFW images on the homepage, one of your differentiators is 'customer support' this really needs to be sorted ASAP.

Even if it's just a static image while you work on moderation, I'd rather not see a disembodied head at 8am whilst eating my frosties..


7 hours ago foxhop just said: "Yeah, sorry about that."

Well sorry about never using your website again. I've just seen things I never want to see again.

Tip: Never launch a site after a day of work. Always launch it when you start your work day. There will always be some problems you like to fix immediately.


Yeah, huge mistake. Sorry.


When I went to this site it showed a thumbnail for YouPorn on the right. Maybe not a good idea to show the most recent thumbnails generated!

Useful service though. I had a need for this a year or so ago.


Yeah, sorry about that. I hope nobody gets flagged for NSFW.

I'll have to rethink the recent image on the right.

If you don't mind me asking, what did you need this service for a year ago? I'm still trying to fit the proper market.


You don't need the most up to date 'recent images'. Toss them in a folder for manual review, and check a couple once an hour, or even less frequently.


Or crowdsource image reviews


I really don't think all that trouble is necessary. A particular user is probably only going to see the homepage once or twice before creating an account, so even just hand-picking 5 images and rotating between them with JavaScript is probably plenty.


I actually built a service like this for internal use a couple of years ago.

I worked at an agency, we were responsible for the web sites of companies of which many were listed at various exchanges. This included handling the publication of press releases. The exchange rules are pretty strict, if a press release is leaked early, you are in a world of hurt. If you publish it too late, you are in a world of hurt. The fines could be draconian in nature.

This led to inane publication work flows, where literally nothing would be allowed to touch anything that could somehow be connected to the internet. The press releases are handed to us in encrypted form. The publication is prepared one day in advance, and at 5am a couple of poor souls have to show up at the office and publish them before the exchange opens, they also have to trigger newsletters. They also had to take print screens of every change they made and file them in a folder.

Because our clients could fine us hundreds of thousands of dollars for mistakes, among other things, I created an automated screen shot service that regularly takes screen shots of certain areas on the client web pages and also downloads the associated html source. External guys audited the application. This is basically just a precaution so that no client can claim we somehow accidentally leaked unwanted information. It's also there to prove that we managed to publish everything in time.

Now of course, once you have something like this in place, someone has to monitor it to make sure it works all the time :) lol...


I thought at first it was some sort of service for porn sites! Porn companies need service providers too...

I think it was for one of my clients that has a directory service for green service providers in Ontario (http://www.greatgreendirectory.ca/Group/Windfall-Ecology-Cen...). For any sort of directory it would be useful to show the website linked to.


I have a porn site as a client, and it would actually be very useful for them.

There are a lot (dozens to hundreds) of generated gallery-style landing pages for affiliates to choose to send traffic to, and having thumbnails of the pages automatically generated would make it much easier for the affiliates to browse through them and pick the ones they want to use.

Out of respect, I'll wait until it doesn't show the latest thumbnails on the homepage before pitching it to my client. :)


I've created my own service like this. The goal was to provide an archive of images for clients, showing their website's content over time and 'proof' of this or that. The setup involved a system to create VNC/xvfb screens, which ran a real browser and created a snapshot further processed with imagemagick. This required a bit of a larger VPS than I would have needed otherwise. There was some overhead and complexity. I researched the systems available at the time, and most of them were either too expensive or inadequate (i.e. only provided thumbnails, and we needed full size snaps).


You could run through the sites you grab the thumbnails from and take an educated guess if they should be shown or not.

Most adult sites have an rta meta tag. You can use that along with Alexa and scraping the keywords + domain name for usual suspect text fragments. Might help block out those horrific images that cant be unseen..

Otherwise, love this service and I'm pretty sure I'll be using it


I got a preview of goatse. This really needs a NSFW tag.


I got the same. Pretty much hit the back button before I even had a chance to check the site out.


It only pulls from your own history, so maybe you need to be careful about what you are browsing!

       /JK


Yeah I got tubgirl.

Welcome to the internet, guys. If you have a public place that displays user generated content, it's going to be flooded with the most vile images around.

Since we've already had goatse and tubgirl, I'm guessing the next is lemon party.


Before reading the comments, I clicked on the link and saw meatspin. I'm glad I work at home, so I can laugh about it instead of being embarrassed.


I saw a picture of an alien and an old man being a lil too friendly. I wish I could laugh but I fear this will haunt me for the rest of the night.

I've decided I'll now read comments first.


Truly sorry.


I see an image of a guy with an inside-out rectum.


I currently use url2png but I could be convinced to switch (and pay).

My biggest need (which no one seems to provide) is support for rendering @font-face web fonts (I use Google's).

If you could get this working, I'd definitely pay for it.


I built http://www.bitpixels.com a while ago, when I needed to generate thumbnail images and the available options were too expensive. It's available for free and serves a few million thumbnails a month.

I didn't add a lot of advanced features (it's a side project), but it's stable and gets the job done.


What makes it better than the alternatives, or (as this is MVP) what is your plan to make it different/better? A company I use (http://xbrow.se/) is relatively feature complete and offers a wide variety of different browser choices, do you plan something similar?


We have huge plans for linkpeek. First step is getting the word out.

I hope to give better support then the other alternatives. I've used a few and found that there was a hole in the market for this type of service.

We want to have a killer api and become integrated into web applications. I want to bring google "instant preview" to anyone who wants it.


minor nitpick: there is a naked fat man covered in poop on your front page.


http://linkpeek.com/screenshot-news-and-updates

LinkPeek.com Apology

Mon Dec 5 16:59:32 EST 2011

I am truly sorry for the grotesque and disgusting imagery posted on our site this morning.

I left the "most recent" image on the website even though multiple people warned me about the possible misuse. I blame only myself.

I lost my lunch today. My nerves are all messed up. I have worked on the LinkPeek software for the past 6 months.

The thought that LinkPeek was used as a weapon against my peers makes me sick.

I feel exploited, I feel embarrassed, I feel ashamed and I have only myself to blame. This is a nightmare and I have no way to fix it.

Yesterday was one of the best days of my life; Today is the worse day of my life.

I don't expect many will read this, but truly I am sorry.

        -Russell


Naah, wasn't that big of a deal. If anything, it was a security lesson for all of us, albeit a very vivid one.


Hi, foxhop.

So there is another service that does this called http://pagepeeker.com -- in what ways are you better and/or different?

Why should I use linkpeek instead?

(And are you worried about using a similar name to them?)


Hash-banged pages are not properly supported, e.g. http://twitter.com/#!/username. It is easy to add - have a look at _escaped_fragment notation.


Thank you for the heads up. We will put some research into hash bang uri support.


I just extracted part of our code to create an open source sinatra app which does the same, you can check it out http://wti.heroku.com/

Oh, check out the screenshot of this very HN thread http://wti.heroku.com/convert?site_url=http%3A%2F%2Fnews.yco... . How meta is that :)


ShrinkTheWeb.com has been around a while and (I think) made a decent small business out of this. However, their site is horrific and it's pretty expensive.

A more recent alternative I've used is url2png.com which has been great.

edit: sorry, not thumbalizr, meant url2png


ShrinkTheWeb was our inspiration for building LinkPeek.com

We would like to correct all the things they do wrong.


I think the first step is correcting the things you're doing wrong, such as not removing the potential for your site to be showing photos of graphic anal sex on the front page. I'm a bit surprised you didn't 1> foresee the potential for abuse or inappropriate images and 2> fix it immediately as soon as it was brought to your attention.

Being on the front page of HN is a grand thing. Unfortunately you may be not only squandering this chance for positive results, but bringing yourself negative publicity.


This is my startup MVP software. Please test it out and let me know what you think.


Do you have 1:1 screenshots available? Being able to generate a website screenshot for specific OS/browser on demand might be a far better niche than trying to compete in a thumbnail segment (even if you will deliver a better service, there's a SEO battle to consider - make sure you know how to win it).


Seems fast and does a good job, but doesn't appear to support https sites?


You are right, I just noticed that too.

After the traffic slows I'll add https support thanks for the heads up! 104 concurrent users right now.


Great concept, but people are taking advantage of the fact that you're showing the latest screen shot on the right hand side of the page. It's pretty grotesque. You may want to fix that.


"Convert Any Webpage to an Image"

Yea, if you mean broken image.

Try any WebGL page (nothing) or any Canvas2D game (loading...) or any page that used 3D CSS (blank)

Links tried: http://mrdoob.github.com/three.js/examples/webgl_lines_color... http://mrdoob.github.com/three.js/examples/canvas_materials_... http://www.htmlfivewow.com/slide41



Have needed a service like this a few times. Current favorite solution is: http://url2png.com/ How do you differentiate LinkPeek?


Linkpeek.com differentiates competition on api speed, customer support, and flexibility.

We plan to price ourselves in reach for both small site (dynamic web designer portfolios) and large link directories link HN. Also a more flexible billing schedule.

Eventually we would like to add features that only our service provides.


Nicely done. I released something similar a month ago: http://instantscreenshot.com

Do you plan on charging for this at some point? I've noticed my bandwidth costs are double due to the traffic coming to my site and the traffic my site generates to get the screenshot.


Along the same lines, url2png has a really nice API: http://url2png.com/


If you don't mind me asking, which stack did you build this in?

I built something identical in about to do the actual HTML --> image conversion. Took <40 LOC iirc to go from url to image.

Always curious to see how many different ways the same problem can be solved. :)


I created something to do this using VNC and Firefox a few years back. At the time, people suggesting using webkit to do a simulated render (which I suppose is what you did) but we required Adobe Flash. So, I created a rotator which could create up to 20 VNC screens on demand, start firefox, and create an image using vncsnapshot.


Neat, vncsnapshot would be a cool way to get output out of multiple browsers too. :)


There might already be services like this, but I'd love something that quickly that will let me see how my page renders in IE 6+7/8 and also runs validation and speed tests. (quickly is the key word)


I was also trying to develop such service but ended up working on something else.

I would sell my domain name ( html2image.com ) to anyone who can make better use of it.


Drop me an email.


I think you can use input type="URL" instead of text to prevent mobile browser autocomplete.


I noticed YouTube videos don't work... webs fonts work so that's good


what does your tech stack look like? how are you generating images?



This looks expensive. How well does this scale?


LinkPeek.com == web scale

It could get expensive but we did hours of research when developing this solution so that it will scale and keep costs down.


Not wanting to be _too_ snarky, but I hardly think "hours of research" translates to "web scale".

Do you _honestly_ think you could keep this up if, say, someone wrote a Wordpress plugin that automatically linked all your eBay sales screenshots into your blog, and Wordpress then installed it on their hosted service?

(Maybe I'm using the term "web scale" wrong, but in my head "web scale" means "the sort of problems Google, Facebook, and maybe Yahoo mostly get right, and almost everybody else mostly gets wrong". And the failure to predict the _first_ thing people would do is point it at goatse is a prime example of "getting it wrong".)


I agree, I was sort of mocking the term 'web scale' as a joke. At this point we wouldn't be able to scale automatically but with some reasonable metrics we could plan capacity accordingly.


How are you planning on pricing this service?


Oh thanks OP. Goatse on the right. Lovely.


Sorry, our users are posting NSFW images.


Heh, and this is surprising why? :-)

It's a nice concept, but isn't it what Silk and OperaMini do? Sort of a pre-render the page and then send that to you? Clearly folks could pull previews off the Google page preview function as well. So I'm wondering at long term utility or a use case.


We generate our own using PhantomJS


Sorry - why is this news again?


Please mark it as NSFW :)


eeeeks. I am not even sure I wanna comment on what I saw.


There are both Firefox and Chromium extensions to do this. What does this service do that they do not?


This sort of service can be automated more easily.


Automatic, self updating, embedable solution.




Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: