Ask HN: Is PDF generation a real pain for developers?

I work in the DOCCM (Document Output for Customer Communications Management) space, which in a certain light is just generating PDFs (well, perhaps a bit more complex than that...). Enterprise deals in this space will have 6 or 7 zeros, so yes, there is a market for generating PDFs. There are already some cloud providers out there (like exari - it's been a while but I remember their stuff being pretty cool.)<p>That's a bit different from vanilla PDF generation, but perhaps a good portion of that is down to marketing. How you position this could be important. You've got your salesforce version - nice :-) That should mean you're selling direct to a business user, so have you worked out what their pain point is? Do they think of it as "generating PDFs" (you've branded it as "Great Reports", so obviously you think they don't!) I don't know anything about report generation on SF, but I'd be shocked if there wasn't already some competition - have you looked in to that? What's your point of difference?<p>Down to your assumptions: 1) I've only ever used FOP &#38; yes it is tedious, but it gives you a lot of power, and that power is going to be hard to replicate.<p>2) Generating large no of PDFs is hard to scale. For a cloud service, though, you need to work out how you're going to deliver a large number of PDFs over the wire. (You'll easily be able to generate more bytes of PDF per $ of CPU than you'll be able to transfer per $ of bandwidth.)<p>3) That sounds like a safe assumption (almost a tautology).<p>4) I'm not so sold on this. If I can get equivalent functionality from a library, I'll use the library. I don't want to add a dependency on an external service (making my own service more fragile) unless it offers a significant benefit. Stripe saves me SO MUCH TROUBLE it's a no brainer. There's is no way I would even think of trying to replicate what they've got. The barriers are significant. What are the barriers to me using a library instead of Pagify? Not nearly as great.<p>Gut feeling, the SF report generation product is much more compelling, but check out the competition in that space.

It is tedious to layout a PDF but actual pain points are handling variable sized images and amounts of text, try formatting recipes to a PDF... Your demo video explains the features but not how it can solve anything that is a major problem to me.<p>The concept here in designing a template to be filled later is a brilliant idea, but I think its the delivery of the product that may pose the biggest problem. PDF generation doesn't have the same level of complexity as video encoding or payments so when it is used, its expected to just work as the environment doesn't change.<p>This introduces what I think is the main issue, if I build a website for a client, as a developer I never want to see it come back with problems, so its not worth introducing the risk of a third party service for the sake of a few hours of tedious work. At the other end of the industry any company producing a webapp may have problems with this simply due to how fundamental the feature can be, PDF generation doesn't require the same infrastructure as video encoding or payments, so they will have very different tradeoffs. Your main problem here is that you have no competition so I have nowhere to go if you close and the design online feature is essentially a vendor lockin.<p>Your assumptions are right but they are what a developer wants, the decisions are mostly made by management where future risk is more important than short term convenience.<p>Personally I think this product in particular would be much more suited to being a standalone package.

Ive used FPDF/FPDI most of the time I used it to generate preformatted mailing labels (the best output method that is compatible with any printer, just format it with scaling "off" in mind.) Maybe a letter or report or two...<p>But the main thing that drew me to PDF utilities was to programmatically fill out pre-existing PDF forms. With FPDI i would load in the form and then overlay the data on top of it. Not as easy as charts and tables, as some data exists as one line per row and other might have several lines per row of data. different positions, etc. (here's a fun form I had worked with: <a href="http://www.cdss.ca.gov/cdssweb/entres/forms/English/CD9600.PDF" rel="nofollow">http://www.cdss.ca.gov/cdssweb/entres/forms/English/CD9600.P...</a> depending of family size or care situation could be multiple pages...) If you can crack stuff like that in an easy-ish to use API, I think you would catch more developers interest.

<p><pre><code> It is tedious to code the layout and design of the PDF using code libraries such as FPDF, Prawn etc. </code></pre> I recently used Prince[1] to generate a PDF version of a web-based book. The PDF needed to be suitable for professional printing (crop marks, bleed, trim, etc.). Prince was the only solution I could find that came even close to being adequate for the job, and it exceeded my expectations. Creating the HTML for Prince wasn't tedious at all. It was almost a pleasure because Prince includes better support for print-related CSS3 properties than any rendering engine I've encountered. There's a web-based service named DocRaptor[2] that is powered by Prince.<p>[1] <a href="http://www.princexml.com/" rel="nofollow">http://www.princexml.com/</a><p>[2] <a href="http://docraptor.com/" rel="nofollow">http://docraptor.com/</a>

The target market that will pay for PDF generation is probably not developers.<p>A competent developer can generate PDF's with the mentioned libraries rather quickly. We deal with a lot of PDF generation where I work, and it is not something that enters our minds to outsource because it's pretty fast and easy to do. For us, the generated PDF's do not need great design or layout, they just need the info in a printable format.<p>That said, you may find a better audience with less tech savvy users that have a lot of reporting needs. I can't really think of a target audience off the top of my head. Good luck to you.

Its a terribly painful. I have tried every single one of the options. I am still finding it extremely difficult to get it working. Once I managed to get it to print to pdf, then started the bigger problem 'Paging'.<p>Paging is quite painful. I am writing my own code to split tables to fit into pages.<p>If you can solve this problem. I would gladly pay.

I rolled up my own solution for rails here.<p><a href="http://ror-snippets.tumblr.com/post/40142053834/converting-html-to-pdf" rel="nofollow">http://ror-snippets.tumblr.com/post/40142053834/converting-h...</a><p>Imo only a subset of people have pdf generation needs (mainly devs)

It's super easy here, I'm generating nice multi-page PDFs with colours and images and all that - from HTML5 based content. I have a similar API available at <a href="http://edoceo.io#pdf" rel="nofollow">http://edoceo.io#pdf</a>

Not amongst the kind of people who read HN. You might get some traction with semi-skilled PHP developers, but they rarely have much of a budget to work with and will generally use whatever is the cheapest acceptable solution.

Can I send a URL to the API and it will return a PDF? That is something I would use for sure.

Browsers let you print a page as PDF. Which use-cases does Pagify solve?

Help us understand PDF generation and its problems.