Status Update 2021

[pointlessone]

This was sitting in my notes for years. I’m still not sure if publishing this is a good idea. It’s bound to stir some discussion and I’m not entirely sure if I’ll handle it right but it seems like now is as good a time as any other.

A bit of history

Prawn is an old gem. It sprang from a very concrete need to generate PDF without external dependencies. I wasn’t around at the very beginning and can’t be certain for sure but this can be pieced together from commit messages, old emails, issues, and PRs.

There was a need (generate PDFs), there were constraints (e.g. no external dependencies), and that’s pretty much it.

The gem grew very organically. That is, it started with minimal features to arrive at the end goal as fast as possible. It’s amazing how much functionality is covered. But at the same time there was not much in the way of architecture.

Later things got untangled a bit here and there but it’s still can’t quite be called an architecture. For example, pdf-core was extracted into a separate gem but it still sticks its fingers deep into Prawn here and there and can’t be used entirely on its own. Likewise, there’s not clear mapping from Prawn to PDF features (or the other way around) which might guide implementation of missing features.

How I became the maintainer

I found Prawn when I needed to generate a few PDF of my own.

It didn’t do all the thing I wanted so I contributed a few features. I liked interactions with the maintainers. I liked giving back to Prawn in return for what it gave me. I fixed a few bugs, improved performance in a few places.

Eventually I’ve got a commit bit. It was Prawn’s policy to be generous with commit bits. It’s still kinda true it’s just, unfortunately, we don’t see many returning contributors to enact the policy.

A bit later then-maintainers decided they couldn’t dedicate much time to Prawn for various reasons and I got an invitation to become a maintainer. For a while I was part of a maintainers team but eventually everyone (but me) took off their “official maintainer” badge.

Current status

In short, it’s complicated. I consider Prawn to be complete and stable. More on this later.

Yet there are a few missing features coming up regularly. For example, support for scripts heavily relying on ligatures or contextual variants (e.g. Arabic or South Asian scripts). Reading/editing PDFs is often requested. Layout tools are lacking. Font support is spotty.

Bugs are aplenty, too. Most are not the severe but some are even if not hit often.

Documentation is not complete and not easy to navigate.

As you can see, there’s much to do.

Hard truth about Open Source repeated time and time again

Stories from maintainers crop up regularly. They’re often filled with bitterness, guilt, regret, defensiveness and sometimes a sprinkle of joy or hope. I’m sure you can recall one.

The latest one is Sorry everybody, I failed with you 😔. I don’t know Pedro but it looks like he’s doing well and I’m happy for him.

I’m doing well, too. I’m not going to complain. But I have to make a confession. An obvious one, too. I was not a very good maintainer.

But it’s not for the reason you might think. Not because there was not many releases recently. Not because bugs are not getting fixed. Not because PRs hang open for a long time.

It’s because I was really bad at managing expectations. And today I’m trying to fix that. My hope is if that gets fixed everything else will become better at least a little bit.

Personal outlook

As I mentioned, I contributed features when Prawn couldn’t do what I wanted it to do. At the moment, it does everything I need it to do. So I don’t work on anything new.

I maintain Prawn because I don’t want to switch to another lib (one may observe, I’m lazy) and there’s no-one else to maintain it for me.

In my eyes Prawn is complete and stable.

Community

I honestly admire Prawn community. It’s not as active as of some other open source projects but it’s still awesome.

I want to brig to your attention a few relatively recent examples that I think should inspire you.

Thomas Leitner (@gettalong) does awesome job triaging issues and answering Discussions. I honestly believe he know Prawns ins and outs much better than I do. And he does this while technically being a competitor working on HexaPDF. I feel a little conflicting by directing you to Prawn’t competition but you should check HexaPDF. It’s awesome.

Cameron Dutro (@camertron) has contributed OpenType Fonts support. It’s incomplete but it’s objectively much, much better than no support. It took a lot of time and a massive effort on his part.

These are only few examples but they illustrate what I think is a great community member.

I want Prawn to be a community project. If you believe there’s a missing feature. Build it. If you find a bug. Fix it. If you see a question you can answer. Go for it. See an issue. Triage it.

The point is build a better Prawn for yourself.

I understand that not everyone can contribute. Some of you don’t have time. Some don’t have the expertise. And I agree that Prawn is not the cleanest codebase out there. No pressure. I hope Prawn is of use for you. But if you can please treat Prawn as your project. Do the work you need done on Prawn. I’ll do my best to not be an obstacle.

I made a call for maintainers a few time since I became a maintainer. I had no responses. Cameron agree do become a co-maintainer of TTFunk after I begged him. There was just massive amounts of code and I needed someone more familiar with it. I’m forever grateful he accepted.

I hired here and there that you don’t have to wait on me to do most of the work on the gems but it probably got lost somewhere on the way or I might’ve been too subtle.

Money

This is not directly related to Prawn specifically but it comes up in relation to OSS and I have some thought and I don’t know when it’s good to stop cutting in editing so I’m keeping it.

So, would money make Prawn active again? I don’t think. There are many reasons. Let’s start with practical ones.

Who’d even get paid?

Maybe Gregory? At the moment he’s the only person financially involved with the project (he generously keep paying for our domain). But he’s not involved with the project otherwise. I personally will be eternally grateful if you helped him recoup the costs. But unfortunately, I don’t believe we’ll be able to muster a sum that would compel him back working actively on Prawn.

Me? Firm no. I can’t claim ownership over much of the existing code. I don’t feel like I’ll rewrite the whole thing to bring much more value. And on top of that I’m in a privileged position where my day job pays way too well to me switch to full time Open Source any time soon if I’m to judge by experiences shared by other OSS maintainers.

Contributors? It’s an interesting though but I have no clue how to make it work from the logistics point of view.

Who’ll pay?

GitHub has Sponsors feature and there are other platforms to collect funds like Liberapay or Patreon. But again, those are usually used for individual to individual support. I don’t think it’s a sustainable model in general and specifically for Prawn.

I know for sure that business use Prawn. Some even for business-critical features. I know that because I myself used Prawn with client projects. It has been mentioned a few times in mailing list. And I’ve got a few private emails starting that.

Unfortunately, it’s not customary for businesses to pay for OSS. I could relicense Prawn to make it easier or more compelling. But I don’t see the point without a plan for spending the funds.

Would that help?

I don’t believe it would. I’m not in it for the money. I don’t believe anyone else is interested in that either. I never saw it coming up in any of the discussions in many years.

I know some of the work has been sponsored in the past. That is, features were developed by an employee for the benefit of the company. It was great for Prawn. I’d be glad hearing from contributors that their time has been sponsored by their employer. But that is external to the project as of now and beyond my control whatsoever.

Hopes and dreams

I’m not abandoning Prawn even if it might’ve looked a bit like that. I hope you (personally) can be more involved with the project in any way you can. I’ll try my best to not be an obstacle. I’ll try a bit to be helpful but just to set expectations right I will not be answering questions at 3am, on weekends, or maybe even in a timely manner.

As I said, Prawn Dows all I need it to do and I’m happy with that.

I hope, you can take on yourself implementing missing features you need for your particular use case. I’ll try helping you where I can.

I hope, you can understand that I will try to avoid any negative emotions. This is an ongoing effort. I haven’t figured it out yet. But to give a glimpse here are a few examples.

I won’t let anyone guilt me into working on Prawn or anything at all. This is specifically to address questions like “When’s the next release?”. It’ll come when it come. Either when I feel like it or when there are enough changes in the master. You can use master in your Gemfile if you absolutely need latest change. You can calmly present arguments to convince me to cut a release. I won’t count it against you. Or questions like “Are there any news on this issue?”. Feel free to pick it up. I might help you just out of gratitude. But guilt won’t get you anywhere.

I won’t tolerate aggressive behaviour in public spaces. I don’t care about your age but I assume you all are adults and know how to be civil. Type it out but take 5 minutes away from the keyboard if you don’t feel perfectly level. Then come back and edit your message. I promise I’ll do my best to assume best possible interpretation but still. Please take into consideration that we’re a global community with different backgrounds and cultures. I also remind that we have Code of Conduct.

And finally, I hope this might generate some conversation. I’ll be happy to answer them.

Disclaimers

I don’t write as much as I should and my editing skills are not good. English is not my native tongue. I’m overwhelmed with feeling and not all of them are good. It’s hot in my office and sun shines directly into my eye through a hole in blind. There’s a pile of email waiting on me when I finish this and I’m not looking forward to it at all. Etc., etc.

[camertron]

Hey @pointlessone,

First of all, a heartfelt thank you for maintaining Prawn and starting this discussion. I know from firsthand experience that maintaining an open-source project - especially a popular one - can be a huge time commitment and source of stress. The Ruby community is as strong as it is because of maintainers like you who give of themselves so tirelessly. At my last company, we used Prawn to generate GDPR portability reports for our European customers. It worked super well.

Thanks also for the shout-out. It was an absolute pleasure working with you on OpenType support in TTFunk. I actually forgot I have commit access haha! I'll try to be more active in managing TTFunk's issues and PRs.

There's absolutely nothing wrong with stepping back from contributing to an OSS project, and for asking for others to help out. The community is here to support you.

As far as money goes, Github Sponsors could be a good option. You might also try reaching out to Ruby Together to see if they would consider helping out with maintenance. They're the custodians of Rubygems, bundler, the Ruby Toolbox, etc. Prawn is well-known and I would even say a critical part of a number of ruby codebases. Ruby Together might be able to provide the stewardship it deserves.

Thanks again for all your hard work!

P.S. Your English is very good :)

[pointlessone]

As I said, I’m not in it for the money. I’ll be glad to pitch someone else to Ruby Together. And I’ll be happy to vouch for anyone who wants to work o Prawn and get paid by them or via GH Sponsors or any other mean.

I’m not stepping away from maintainership. Not really. I do t expect spending less time on Prawn than I currently do. If anything I stepped away years ago. I’m still here. I’m just trying to set expectations right.

PS: Thank you. 🙂

[gettalong]

Incidentally, I recently did some very small paid work for a company that uses Prawn and it seemed that they were happy someone was available to do that work. Probably because understanding the code base is one thing but without the necessary PDF knowledge it is hard to make some changes or find/fix bugs.

And I also look at prawn/pdf-core/ttfunk regularly as part of me benchmarking HexaPDF and try to fix things there as they pop up. So that may count as a (very small) company doing work for Prawn 😁

[pointlessone]

@gettalong That's awesome! Thank you for sharing.

[gettalong]

I think you are doing well! Thanks for maintaining Prawn!

Maintaining something popular sometimes requires hard decisions and the best you can do is to do it the way you feel good afterwards. When releasing kramdown 2.0 I yanked out everything not strictly in the "core" into extra (unmaintained) gems so that maintenance would get it easier. It was a hard decision for me but ultimately the right one. If these extra gems are interesting for some, a maintainer will be found (as was the case with one or two of them).

And thanks for mentioning me! 😁 I do what I can/have time for.

One thing I would do to make things easier to maintain would be to kill off pdf-core and integrate it back into Prawn. It can't really be used as a standalone gem and as far as I know is only used by Prawn itself. The split between pdf-core and Prawn also seems rather artificial to me and, in my opinion, just complicates things (e.g. when trying to triage a bug).

[pointlessone]

Prawn/pdf-core split is artificial. pdf-core was extracted from Prawn. Unfortunately, it wasn't quite completed. And the lack of any proper architecture made the final separation rather difficult. I'm on the fence about it. On one had I like the idea. Untangling different parts of Prawn might guide it to some kind of structure in the codebase. On the other, it's a lot of work. I flip between the opinions like every season. So… Maybe sometime one of these will materialize.

[artbot1]

why not substitute prawn/pdf-core with hexapdf, which doesn't seem to have the same architectural flaws prawn seems to have (never looked into hexapdf internals, prawn doesn't seem that bad by the way) ...

starting a kickstarter or whatever campaign for hexapdf to change AGPL to MIT license and everybody(?) could be happy ... (take blender as an example)

@gettalong: have you ever thought about it?

[pointlessone]

Rails is a rather weak argument for MIT. Rails is essentially a consortium project. Most of the core team is employed to work on Rails. Other contributions are motivated because Rails is a critical piece of infrastructure for them. So critical that it's easier/cheaper to contribute to the framework rather than migrate to something else.

"Perfect PDF support" is an amazing value proposition. And I'm sure any PDF support is critical to at least some businesses.

Let's take Prawn as an example. It was (and arguably is still for many use cases) the best PDF tool for Ruby. It was out there for over a decade. It it liberally licensed. In addition to GPL there's Prawn license which is even more free than MIT. And somehow financial support has never matrialized. Nor did other contributions coming from caring businesses.

HexaPDF (and Prawn, for that matter) is available for anyone to use. How would MIT help improving that? HexaPDF has a legit financing option. It's not even expensive. Single MIT would cut that. That would benefit a lot of people (namely, freeriding businesses) but it would be to the detriment of the only person actually creating value here.

I'm not bashing MIT. It has it's place. But MIT is far from universally good. It's not even mostly good. I'm saying that from the point of view of sustainability. If you put together something in a weekend and just want to share it with the world then yes, MIT is OK. But if you have any long-term plans with the project and want to make it sustainable, MIT is probably the worst option.

---

This is sort of an off topic/tangentially related discussion but I'll allow it. If you want to argue the benefits of MIT go for it but please make an effort beyond abstract slogans and add at least some argumentation.

[artbot1]

i don't want to argue about the MIT license here, i use prawn a lot (also because of it's license) and i like it (also your work and of the original maintainers)

just imagine(d) hexapdf (or however refactored prawn) being sort of
actionPDF (or whatever basecamp would make it sound in their marketing newspeak :-),
then it would be "critical piece of infrastructure" for a lof of people ... and maintaining it would maybe rest on more shoulders than yours ...

[gettalong]

@artbot1 So the dual-licensing thing for HexaPDF is actually a very deliberate choice since it was clear for me that I could only justify putting so much effort and time into the project if there was a financial component. Waiting for companies to support/donate to an MIT licensed project can be a long wait...

I have considered making a compatibility layer for HexaPDF based on Prawn's API but the way both projects do things is so different that supporting the Prawn API would be quite an undertaking. Since it is possible to do that without changing HexaPDF's core, anyone could start a separate gem implementing that interface.

[artbot1]

@gettalong: i strongly support your view about the financial component. this was also my suggestion:
start a kickstarter campaign for an amount you could be (more than :-) happy with. if the campaign fails, everything remains as it is now. if it was successful, then hexaPDF would get an additional "MIT license" like license (commercial licenses with support etc. could still be an option)

[pointlessone]

@artbot1 I can not help but wonder why do you like MIT license so much?

[strzibny]

Thanks for the update and your work on Prawn. Without people like you, https://github.com/strzibny/invoice_printer wouldn't be possible to make (it's just a little layer on top of Prawn).

[pointlessone]

I'm glad people find Prawn useful.

[yob]

Thanks for the thoughtful update ❤️

I was one of the original contributors. I'm still lurking around, but I'm not doing any PDF work in my day job so haven't had time to contribute to prawn in a long time. (git says my last contribution was 8 years ago, and my eldest child is 7.5 years old. Co-incidence?). A little bit of my spare time goes into keeping pdf-reader on life support, which keeps me busy enough.

One of the original goals of prawn was to keep everything pure ruby with managed objects. At the time, the other options (html to pdf via princexml or wkhtmltopdf, libpoppler via the Ruby/GDK bindings, etc) all had some nice features but awkward tradeoffs. There was and remains a large class of PDF generation needs that fit a different set of tradeoffs where pure ruby is a great solution. I'm stoked to see prawn is still around and being kept relevant.

If it helped I'd be happy to chip in a bit via GitHub Sponsors to help keep up the maintenance. It sounds like you're not keen to go that way though. You have my ongoing thanks in any case ❤️

[pointlessone]

If I were to describe my experience maintaining Prawn I'd say I'm still around because I sand on the shoulders of giants, one of which is you. Thank you for all the work you did on Prawn and pdf-reader. I'm glad you dropped by.

Money is not the driving force or a way to improve my involvement. And I'm not the one spending money on Prawn. If you must, ask Gregory how to pass a few bucks because he's the one paying for our domain.

[petergoldstein]

@pointlessone I'm an experienced open source maintainer who has experience maintaining code bases initially developed by others (e.g. https://github.com/petergoldstein/dalli). As such I understand the challenges of this kind of maintenance role. Your work is really appreciated.

I don't use Prawn particularly frequently, but I'd be happy to step into a maintainer role on an interim or a more permanent basis. I have some time at the moment so I could help triage the issues and PRs, shepherd what makes sense to commit, and release updated versions. Please let me know if that's of interest.