Is ZipTie a good candidate for a fork?
First, I'll give you some background. The project was created in November of 2006 and was sponsored by a company called Alterpoint. This was great because they funded four developers to work on ZipTie for over two years. Here is a link to my first review of ZipTie:
I started using ZipTie on a university network that contained hundreds of network devices. ZipTie automatically discovered my devices by crawling the network. Then it automatically made a backup of my configs, and kept a history of revisions that I can use to recover from. Additionally, I can send commands to hundreds of devices at once, saving me countless hours of time. The amazing thing about ZipTie was that I could accomplish all this with just a few clicks using the well-designed graphical interface.
Things were going great, and ZipTie continued to improve by adding more features, support for more types of equipment, and better stability. ZipTie also created the ZipForge, which is a place for network engineers to upload scripts to perform specific actions or to support additional device types.
Problems started to appear when the company sponsoring ZipTie, Alterpoint, decided to change the project from the Mozilla Public License (MPL) to a proprietary license that they created. At the same time, they changed the name from ZipTie to Network Authority Inventory (Yes, from two syllables to ten!). You can read my opinion of these changes on a previous post.
The most recent news about ZipTie is that the entire development team was laid-off, and Alterpoint was purchased by Versata this spring. Since that time, there has been no mention of what will happen to ZipTie. Judging by the fact that they fired the development team, it doesn't look good.
Is ZipTie a good candidate for a fork?
When discussing the benefits of open source software, it is often mentioned that the code can continue to live on, even after the sponsoring company fades away. Is this possible with ZipTie? In my opinion the value of the code is very high, so it would be a shame to see it simply wither on the vine. So what will it take for ZipTie to live on as a new project?
Does the MPL license allow ZipTie to be forked? If so, what is required? I'm assuming that we may need to rename the project, since the name is probably trademarked. Here are some possible names that I have come up with:
NIZ = NIZ Isn't Ziptie
NINEZ = NINEZ Is Not Exactly Ziptie
PUNT = Powerful Uncomplicated Network Toolkit
VUNE = Virtual Unpaid Network Engineer
Please provide feedback on what the next steps are in this process. I am not a developer, so I will not be able to contribute progamming skills, but I bet that there are programmers who are interested in working on this project. If you are interested in helping with this project in any role, please post a comment.
Thanks Interweb!
Two cents, two sentences.
ReplyDeleteFirst rule of forking: you do not talk about forking if you are not a developer.
Seriously, my advice: pick up an editor and start forking or rethink this through.
Anonymous - Thanks for the input. I have to disagree with your comment that only programmers should allowed to talk about the future of an open source project.
ReplyDeleteThat would be like saying, only the people acting in the Christmas play can talk about the production. Everyone else who contributes in other ways have no right to talk about the play.
What about the people singing in the chorus, the musicians in the orchestra, the director, the person running the lighting, the sound guy, the choreographer, the person who gets drinks and snacks for everyone, the person who babysits the actor's kids, the person who makes costumes, the person who puts up posters about the play, the person who helps decorate the stage, the person who makes programs to hand out to the audience, and the ushers who help people find their seat. Each of these people have important roles in the success of the Christmas play, even though they are never seen on-stage.
In the same way, anyone in an open source community has a right to care about the success of the project. Ziptie is like an undiscovered gem, and I feel that it is my job to make sure that people who can benefit from it are aware of what it can do. Once enough people learn about the project, ZipTie can start to become a self-sufficient community.
I want to further explain the roles I have had with ZipTie.
ReplyDeleteAdvocate/Champion = Once I discovered how useful ZipTie was, I wanted to share it with other people. I have done so on this blog, I have presented ZipTie to over a hundred network engineers at a conference, I have provided quotes for Alterpoint marketing materials, I have done an interview with Network World, and I have shared my ZipTie experience on several mailings when people ask about network management tools.
Beta tester/Bug finder = I have deployed ZipTie on one of the largest production networks it has run on. As such, I have run into many bugs during the development of this young project. I have done a good job of documenting these bug using the bugzilla tracking tools that ZipTie developers deployed.
Feature requests = Similar to finding bugs, I have identified several areas that ZipTie developers could change to improve the user experience.
Do you think these roles were useful? Do they give me the right to talk about forking ZipTie? If I don't talk about this, who will? Will the thousands of man hours of programming time go to waste and the ZipTie project slowly fade away? I surely hope not, and I am doing my best to prevent that from happening.
Tristan,
ReplyDeletePlease don't take anonymous comments too seriously. You certainly have the right to suggest a fork, regardless of whether you're developer or not. Not only Ziptie is no longer open source, the project essentially seems to have been left to die by the company.
Having said that, I have doubts it can be revived by a fork. Why did the original project failed? Were there a lot of (any?) contributors to the project outside Alterpoint resources?
The way I see it, the failure of this project does not bode well for role of open source in IT management.
As far as I could see, Alterpoint did a lot of things right. They had high quality developers, permissive open source license, and genuine effort to make it easier for other companies and individuals to participate. The project had backing of a corporate entity that provided support and descent pricing compared to proprietary alternatives. Their technology choices were excellent and product solves an important and difficult technical problem. So what happened?
Unfortunately, there was no post-mortem by the folks who were involved to have these questions answered. (some of them were required not to comment?) But the answers to these questions are key to determine whether the project can be revived (forked).
As far as I can see, one problem is most users of this project are network admins, and network admins are not developers. As you've mentioned above, admins are very valuable and can contribute in many ways, but who will develop the software? Many people can contribute but there needs to be some core developers who dedicate significant amount of time. What is the model for these developers to get paid? Can another commercial entity succeed where Alterpoint could not?
Large corporate customers can get tremendous benefits by investing resources in such open source projects. Maintenance costs alone for proprietary software products similar to Ziptie can be in 10s/100s of thousands of dollars. Why not dedicate a developer to this open source project instead? Unfortunately, this model has not taken roots and enterprises have little to no participation in open source projects.
How many users did the Ziptie project have? Who are (were) the largest users? Can they be persuaded to commit or fund resources to the project if there was a fork?
My guess is if there were individuals willing to dedicate their own time, this would have already happened. So the only alternative is for there to be a new model where quality developers (like the ones who have developed this product originally) can be funded.
Finding a new model requires more than development skills hence your initiative is more than welcome :) Good Luck!
Berkay
Yes, I think ZipTie/NAI might be a good candidate for a fork.
ReplyDeleteFirst a positioning statement...I work for one of those "proprietary software products"...which is why I chose anonymous...not sure how my employer would feel about this post. And no I'm not the first anonymous above with the weird requirement that only developers comment.
ReplyDeleteAlterpoint was a strong competitor for some time and to be honest I'm sad to see them go the way they did (of course I'm glad it wasn't us :)).
We actually considered open sourcing our driver infrastructure the way they did (before they did it of course, wouldn't make sense to try to have two standards...). In our analysis we found that most of our customers wanted us to do the work and were not interested in writing their own drivers. Indeed they were using our software to reallocate resources, not invest in more tools. Much of the user base for this type of thing is network admins, not hardcore coders. That's why much of the contributions to ZipTie came from Alterpoint themselves. The concept was good, but just not a big enough base to really sustain it. I think unless you can find a group of either interested college students, or another company who wants to make a run with it...it will be hard to revive this.
Anyway, I'm personally torn on this one...I firmly support open source, but this one competes with my livelihood :)
I do wish the best of luck to you though.
If the original authors changed the name, did they abandon it? Could you do something like ZipTieNG (next generation) to show its link to the old project?
ReplyDeleteYes, MPL allows forking.
ReplyDeleteI guess the first thing that's required is a copy of the source code. Can't just get the latest from ziptie.org revision control, since that version is not MPL.
CVS has an entry in changelog "update license text" 29 Oct 2008 - This may or may not be the cutoff.
I am the anonymous who posted the first comment.
ReplyDeleteI still maintain my position. Berkay explained above the general reasons quite well.
Forking a project is a lot of work. Often it involves more work than writing the whole thing from scratch. First and foremost, forking needs dedicated developers. Endless man hours of labor.
This is the reason why only those who are willing to do the work should talk about possible forks. While this does not apply here since the original project is dead, it is always good to remember that shouting "fork, fork!" in the internet is often considered rude, stupid, and abusive. It is better to shout only after you have something to show.
No offense; you seem to be genuinely interested in the project. Good luck.
From this post it seems like it's possible:
ReplyDeletehttp://inventory.alterpoint.com/forum/viewtopic.php?f=3&t=150
Thanks Tristan. Thought I'd let you know that your support is appreciated and that there are still people who care.
ReplyDeleteOctober approaches.
It seems a shame. I've been playing with NetAuthority Inventory for the past couple of years. I was very excited during the NEMO beta period; only to find out when the final version came out - the product was changed. Now its been nearly a year, without updates.
ReplyDeleteWhy put all the effort into developing a nice web base interface for everyone, and then basically abandon?
Ok - so I was on the phone with a NetAuthority product meeting - here is the sum-up regarding NetA as a whole, both development, and the Inventory product - Per their comments: "There was a staff turnover, but some of the best developers who have been involved, are still around." Versata, they say has improved their product line - they went from 12 dedicated developers to a pool of 1200 in India.
ReplyDeleteThey also re-affirmed their commitment to the open source software; mentioning it has not been forgotten. Good news - lets see if they keep their word.
The inventory.alterpoint.com website seems to have gone off air. Has this been moved to a new URL?
ReplyDeleteDoes any one know where to get the source code for NAI/ZipTie version 2008.10? The cvs.ziptie.org and fisheye.ziptie.org sites are inactive.
ReplyDeletethe site's responding now, probably a web server hiccup.
ReplyDeleteAnyone knows where the ziptie source can be found on the net? The last version under Mozilla license? Cant even find any mirrors of it,,,
ReplyDeleteThe last couple of days I've been looking for Is ZipTie a good candidate for a fork and gladly I stumble into your blog, it has great info on what I'm looking and is going to be quite useful for the paper I'm working on.
ReplyDeleteBTW is crazy how many generic viagra blogs I manage to dodge in order to get the right site and the right information.
I would not be too trusting of what Versata says (I used to work for a company that got bought out by them). One minute they were acting like nothing was going to change and it was going to be business as usual, the next minute everyone is being laid off. They are a viral patent driven company. They make the majority of their money by suing other companies over "patent infringement". I would like to get my hands on the source code for 2008.10 as well. If I manage to dig it up from somewhere I'll post a link to it here.
ReplyDeleteDoes someone have the sources of the Project anyway?
ReplyDeleteIt seems that it cannot be found on original sites?
I would like to give it a try?!
Anyway is someone monitoring the comments it looks like there are few "spam"'s
Where is ZipTie source? The website is long gone?
ReplyDeleteComing soon... http://code.google.com/p/ziptie-ng/
ReplyDeleteDaniel - Thanks for posting the link to your planned project to breathe new life into ZipTie.
ReplyDeleteI know that some other developers have expressed interest in a fork, so hopefully you can coordinate your efforts with them.
I've put together a fork of ZipTie under the name Xerela. The project is entirely rebranded because AlterPoint still owns the ZipTie trademark.
ReplyDeleteThe project is at http://code.google.com/p/xerela/
You can download a Windows installer at http://code.google.com/p/xerela/downloads/list
With installation instructions at http://code.google.com/p/xerela/wiki/Installation
I was a developer at AlterPoint in the days before ZipTie and a consultant in the days after the license got changed, and I'm hoping that some of the original ZipTie developers will want to get involved now that most -- if not all -- of them are free of any non-compete agreement with AlterPoint. Since there's no longer a corporate sponsor supplying full-time people, this will truly have to be a community effort for it to work.
If you're interested in getting involved, you can reach me at aurelio@xerela.com. You don't have to be a developer -- in fact, one of the biggest needs right now is help with documentation. While the ZipTie source was under the MPL, it's not clear to me that the same is true of the documentation.
Also -- Daniel and I are working together on this, so Xerela and ziptie-ng are not competing projects.
Aurelio, I hope that you can find interest from the other former developers because this software would be a godsend to me at the moment. I am trying out Xerela now and hope you can maintain momentum.
ReplyDeleteDears, Whats new with the Ziptie/Ziptie-NG/Xerela projects? Is it in full swing?
ReplyDeleteAurelio, I've noticed that you're only focusing on windows based installers for Xerela. I'll tell you right now that it's going to gain absolutely no traction unless you continue to supply Linux compatible packages as well.
ReplyDeleteAs per "Anonymous", there isnt much of an open source coder community in windows compared to linux.
ReplyDeleteIt would also be worth moving the project to github (or git in general) to make contributing as easy as possible.