Valentina Fork Settles Down As Seamly2D, Valentina Goes On

Valentina Fork Settles Down As Seamly2D, Valentina Goes On

Four months into a bizarre fork of Valentina, free pattern-making software for fashion designers, Susan Spencer's leg of the fork finally gets rebranded as Seamly2D.

There are now two projects that share the proverbial 99% of code base: 1) original Valentina project, forked by its founder Roman Telezhinsky, 2) Seamly2D, managed by Valentina's other founder, Susan Spencer. But let's roll it back a bit.

The Story

The project was started by Roman Telezhinsky (Ukraine) and Susan Spencer (USA) in 2013. Both founders had previous attempts at writing software for pattern-as-in-clothes design. However, within the Valentina project, Roman took the role of writing the code, while Susan quickly gravitated towards community building, PR, handling financials (paying Roman's salary, in fact) etc.

Early on, Roman took position that basically boils down to this (opennet.ru, 2013):

I work on this project for myself. If anybody else needs it—great. If nobody else needs it, it's fine as well.

Depending of where you are coming from, this either contradicts or complements his more official statement (Valentina blog, 2013):

It's clear that a single person cannot realistically create such a program. So I made it an open project, hoping that I won't be the only one interested in it. I hope it doesn't stop at that.

Despite this rather blunt classic approach to publishing software under terms of GPL, users soon started gathering around the Valentina project. The two main reasons for that were technical excellence of the software (despite a lot of rough edges) and solid community work.

The former can be explained by introduction of parametric design to software for end-users, which greatly simplified making adjustments, as well as refitting an existing design to a completely different person.

Moreover, with over 50 pattern-making systems supported, this made the project somewhat popular with designers of contemporary clothes as well as the historical recreation community, since a significant part of the supported systems cover Victorian tailoring, as well as garment cutting from even earlier centuries.

There's something else that should be factored in to explain public's interest in Valentina/Seamly2D. Pattern-making software is mostly proprietary and very expensive even for personal use. Top-notch systems like Gerber AccuMark and Lectra Fashion PLM are targeted at large companies and are in the general arm/leg/kidney ballpark price-wise. If you know exactly how much either of them costs, congratulations—you are an owner of a large fashion business with hundreds of employees.

Less expensive options typically start around $1,000. Some cheaper (and simplistic) solutions exist, and even then vendors would try to charge you for every single extra feature.

And, to the best of our knowledge, none of the above have native Linux versions. Needless to say, none of them is free-as-in-speech.

A user who commented on sodaCAD blog back in 2014 pretty much nailed it:

I've been in the pattern making industry for over 20 years and we REALLY NEED a free/cheap/open solution. It's almost impossible to hire skilled operators in New Zealand simply because nobody can afford to buy the software and get skilled up in it.

That's why breaking the Valentina team in two was dangerous, if inevitable. But this is not the usual "a couple of programmers had a technical argument". Digging into the story of the conflict between the founders has been an exceptional, if frightful source of insights into the world of Things That Can Go Wrong On So Many Levels.

  • Language barrier? Check.
  • Mutual misunderstandings and apparent lack of persistence to clear things up? Check.
  • Huge project vision clashes? Check.
  • Being borderline rude to potential contributors? Check.
  • Alleged locking one founder out of direct communication to potential partners by another founder? Check.
  • Social awkwardness of one founder enabled by the tendency of the other founder to sweep the dust under the rug? Check.

Arguably, so far the most sensible comment on the whole situation comes from Mario Behling who, at some point in the past, unsuccessfully tried bringing Roman to live in Berlin and work on the project in a hackerspace:

In my opinion they should just calmly do their own things and let it be. I think their worlds are just too far apart.

It's hard to tell how calm they can get. In his most recent post, Roman summarizes his vision of working with a community and uses what one might call "brutal honesty". The statements go well into the uneasy territory, breaking almost every rule of contemporary community management. If anything, they hint at exactly how difficult it could be working with him for other contributors—something he readily admits in both private conversations and earlier public posts.

And then what?

We could leave it at that, was it not for the fact that four months into the fork, the amount of confusion about the two separated projects is still staggering. Not in the least place because it's caused by actual stakeholders.

Case in point. A few weeks ago, Susan Spencer launched the Fashion Freedom Initiative (FFI) which is:

...an open community of indie designers, forward thinking businesses, artisan producers, makers, crafters, hackers and doers. We are working together to build and run our own, independent chains for global fashion production.

The initiative seems like an interesting approach to solving quite a few things that are wrong with the fashion industry. The founders appear to rely on Seamly2D as its strongest community-building tool. So it's expected that the project started posting user stories.

The first such story, a Seamly2D testimonial by Megan Rhinehart, founder of Zuit, is a great inspirational read, save for several statements she made.

One of the things I love about Seamly2D is that it is getting translated into so many languages.

It's not. The Transifex account that Susan Spencer keeps pointing users to is owned by Roman Telezhinsky. They are not translating Seamly2D. They are translating Valentina and probably don't even know that.

Moreover, she couldn't be using Seamly2D, unless it was a private build from Git master made within the last couple of weeks. There are simply no builds of Seamly2D to download yet, nor have there been releases of Seamly2D. The 0.6.0.1 release was made a full month prior to the final rebranding. Susan Spencer got the valentina-project.org domain name and the website as part of the separation deal. The downloads section of the website still distributes Valentina builds. It even says "Valentina" right on the front page, next to "Seamly2D".

[Seamly2D] is cloud-based so I can see what the tailor sees. I could potentially add users to help with pattern design and quality control.

Seamly2D is not cloud-based, nor is Valentina. It's a Qt/C++ desktop application that has to be downloaded and installed. When asked for clarification, Ms. Rhinehart replied that there was "a third party app to run it on the cloud" involved. As of December 7, the testimonial retains the original, unedited statement.

It also doesn't help that Seamly2D has two simultaneously maintained GitHub repositories (more on that later). Some of that confusion can be explained by the fact that the separation agreement was made hastily, angry conversations went on for a while, and there were no clean cuts.

Present State of Affairs

In terms of writing actual code, this is what things look like at the moment.

Roman more or less maintains the programming pace, fixing bugs, making various enhancements, writing new features, and publishing test builds. August through October was a busy time for him, less so for November, and he expects December to be a slow month for the project as well.

Code-wise, Seamly2D isn't as efficient so far. Currently, the project confusingly operates on two GitHub repositories:

  1. The one with rebranded repo name and all (or most) converted issues from the Bitbucket tracker, yet without latest changes: https://github.com/fashionfreedom/seamly2d.
  2. The one with the old name and yet with all recent changes, including rebranding in the source code and visuals: https://github.com/valentina-project/vpo2

In fact, since August, changes in what is now Seamly2D code base boil down to rebranding, updating/fixing the build system and setting up automatic builds on a new account, updating various build/contribution related docs, renaming icons, and improvements in generating tiled PDF files. That is, the vast majority of changes doesn't fix bugs or introduce new features.

During a conversation on September 11, Susan Spencer stated:

Since Roman left, we've received offers to contribute from four programmers. They are waiting on the issues list to be recreated.

This is an important part, because the alleged pushing away of contributors by Roman was one of the biggest concerns mentioned by Susan.

However, three weeks after this step was completed, source code changes still weren't pouring into the repository. We asked Mrs. Spencer for an insight on that, and then a weird thing happened:

  1. She changed the narrative into what boils down to "we do have programmers, but they are currently unavailable".
  2. She then provided a rather believable explanation for each "missing programmer" case, without naming anyone or giving away too many details in order to protect the privacy of the alleged future contributors.
  3. Following that, she mentioned another technical detail about all of them that, if published, would raise questions about possibility of actual programming to be done in the project.
  4. Finally, she specifically forbid publicly mentioning specific information she provided out of "fear that ... there could be a big questionmark on our community" within this article.

Instead, Mrs. Spencer provided this statement:

I would like for the take-away from all this to be that our all-volunteer community is handling the situation rather well. They are an open, honest, and upstanding group of nice people who care about each other and about the project. I'm quite proud of them.

All in all, Susan Spencer seems genuinely defensive of the community she helped growing, although in this particular case this leads to questionable PR tactics.

Aftermath

It would be extremely easy to blame either side for what's going on with both projects currently. However even from what's left unmoderated in the forum it's clear that there has been a lot of mutual hostility, but above all—lack of understanding coming from both founders and community members. Some of it continues to pour out one way or another.

Maintainers of both Seamly2D and Valentina emphasize that their projects are doing well. However the former has been mostly lacking visibly active developers since day one, and the latter doesn't get nearly as much community awareness as before.

In the coming months/years, we are likely to see for ourselves, whether a community/PR manager can build a team of developers, and whether a developer can succeed in building a strong dedicated community.

If you ask, which project you should be tracking from now on, the best we can get you is "both, if you can stand occasional passive aggression outbreaks and nasty remarks". Nobody actually promised that free software would be a peaceful ecosystem. But it will get better.

Was it useful? There's more:

10 Responses. Comments closed for this entry.

  1. It is so sad to see this happening. Valentina/Seamly2d is such a needed tools in the open source community and such personal clash will negatively affect the whole community taking sides and splitting the effort. Hopefully we can see a creative resolution ~ I feel for both parties here!

  2. Hi, every one I strongly agreed with Gef, not only Valentina, all open source application should not be close at any cost. This news is very painful for open community.

  3. Roman Telezhynskyi 11 December 2017 at 12:30 pm

    Alexandr, thank you for the article, very honest.

  4. I hope good things will come from this turbulence. For those of you who are willing to explore alternatives, they’re also https://freesewing.org.

    Full disclosure: I am the project maintainer

  5. I’m the developer at sodaCad and I nearly predicted this very situation when Valentina was first getting traction. There is a lack of focus in the project. Who are they developing for? When asked they say everybody. That doesn’t work in the real world as professional pattern makers work very differently from the way Valentina does. My own project targets those people working in the industry as professional pattern makers. The user base is small, but in great need. sodaCad is stalled because I lack the time and advanced programming skills to take it to the next level. I haven’t give up though.

    BTW, there is a Linux compatible commercial CAD software for apparel pattern drafting. It is http://www.padsystem.com/enPAD system. I downloaded and tested the trial version. It does work fairly well, but the user interface and keyboard shortcuts were not to my liking, not to mention cost is still a factor. There are three CAD systems available that average $10,000. You can do an outright purchase but many of them now prefer a monthly rental.

  6. Esther, Valentina/Seamly2D compare to your project are super healthy because of constant grow. Project as leaving being, if it stops it dies. It is good that you haven’t give up, we need your project to continue live. I personally many times said people about your project. In my eyes we are not competitors, more like colleges. We share same area, but trying to fix issue in different way. So, don’t be afraid traction around Valentina/Seamly2D.

    I know your point of view about what is right and what is wrong. But as an author you should understand difference between theory and practise. Instead of talking we should better to concentrate on our projects. And by success of our projects will prove that we are right.

  7. I’ve been a patternmaker going on 4 decades. I own an arm/leg/kidney CAD system (StyleCAD @$9K) and I don’t have hundreds of employees. In fact, when I first got StyleCAD, I had no employees. Today, I own a small but high end sewing factory as well.

    This is the point that no one (but Esther) seems to understand; any opensource CAD system is going to have to interface well with existing CAD systems and the former will need to provide the same functions as the commercial versions. Employee headcount and company size is immaterial. The functional attributes of the software to fit the job requirements is critical; without a match to industry tasks, OS CAD is nothing but a interactive computer game.

    To wit, say a “designer” makes a pattern with the idea of having a sewn product or clothing item produced in quantity. If the designer owns a factory then nothing more need be said. If he/she doesn’t have one and needs to hire a factory like mine, how do you all think that the pattern he/she provides, will dovetail neatly into the production process? Currently, it can’t, not with OS CAD. Two things leave me gobsmacked:
    1. the blind naivete about how the production process works, and
    2. what appears to be a complete absence of curiosity about the production process.

    Or perhaps one thinks they’ve got enough, and good information… this is terrifying. Other than Esther, I don’t see anyone affiliated with this project, in any capacity, as being qualified to speak to any of this. People who can, are too busy doing it for a living. I can say that some of us who do this for a living were excited about this project but we exited way back when we saw that the development team had zero priority for addressing the commercial needs (in any meaningful way) of independent designers.

    For my part, I was excited when this project started up because it could help my customers save some money but I saw that nobody really cared about the most salient needs we have in a CAD solution.  Meaning, Valentina/Seamly will not help indie designers get off the ground which leaves it in the realm of hobbyists. If OS CAD doesn’t add industry compatible tools that sewing factories use as a matter of course, I don’t understand why people even want a CAD tool as they could just as easily (or more easily) use a pencil and paper.

    I don’t know why I’m even posting this except that I’m an eternal optimist and hope that somebody will listen, and develop a tool that can help designers get their businesses off the ground in a more cost effective way. For my part, I can’t use these files. They won’t import into my CAD system so I can’t grade, can’t calculate yields, conduct cut order planning, make markers, none of that.

  8. Alexandre Prokoudine 13 December 2017 at 7:21 pm

    @Kathleen,

    From your comment it’s unclear if you merely followed the project or if you communicated your interests to the team. As someone who isn’t professional in fashion design/manufacturing at all, I’m also unsure what file formats exactly you would expect Valentina/Seamly2D to export to.

    Oh, and you should be reading that part about hundreds of employees with a winky face. Not everything is meant 100% seriously. Some things are merely jokes. This was one of them.

  9. Kathleen, i don’t know who you are because cannot recognize your nickname. But the fact that you made so big comment give me a hope.

    In fact many misunderstand how they should participate with the project. You made the same mistake as many before you. It is not enough to just come, say hi, i have a business and i want this for my business. And when you hear NO ... we all know what happens next. See your comment above. You did not try to help this project, you come to satisfy your needs. Satisfying needs it is the reason why we all here. It’s normal. But we do it in different way. You cannot understand one simple truth.

    Why do we all here? Because ourselves we cannot get what we need. But together we have a hope. It is about sharing solutions. Putting them in one basket and it becomes full. I want to see a project where strong people participate to become even more stronger.

    But from what i see around me only few really understand what it’s really mean to build this community. It is like with medecine. People complains that it cannot cure them, but at the same time don’t want spend their lives to resolve this problem. Your approach will work only for proprietary solutions. Those who unwilling to take their lives in their hands will always unsatisfied. You did not propose anything for this project except of your needs, this is not enough to get what you want.

  10. And who said you that Valentina is closed? This is issue with community that mostly is not capable to help.