9 ms·

Show HN: Yazz Pilot – Self Service Apps Without the IT Department

95 points by zubairq 7y ago

zubairq 7y ago
Hi Zubair here, I just posted a project called Pilot that I have been working with on and off for many years now. The tool is supposed to be a Visual basic style drag and drop GUI builder for enterprise apps. the interesting part of the backstory is that this is my 7th attempt to build this! So the back story is: I first thought about building a web based development tool since I was at University where I built a Hypercard style development tool called OpenPage (using C++ on Solaris) as my Computer Science project which scored as one of the highest projects of the year at Machester University, UK (1994). My beginners confidence didn't carry on into anything concrete though, until 2000, when I wrote the second version in Java, but this didn't go anywhere. For a third version I tried to develop this in Javascript, Ruby and Sintra in 2007, but I didn't realise how difficult a problem this was. Also, in 2008 a fourth attempt was made where I tried again in Erlang, but it didn't go anywhere. So I hung up my boots and then in 2013 after building a startup NemCV with Franco Soldera I tried a fifth time and built a Clojure/Clojurescript based tool called Clojure on Coils, which was pretty cool, but most people thought it was a joke. Eventually after a couple of years I realised that I didn't understand the problem well enough, so for attempt six decided to outsource the problem by investing in a project by Chris Granger and Rob Attori, called LightTable/Eve. They tried to solve the problem of building a usable development system, but by 2018, they too realised that this was a huge task, and very hard to monetise. So, I took a step back and realised that I now had a lot of knowledge about the subject, but needed to find a way to build something that can be commercially viable as an entry point, as this is a problem that 10,000s of companies have tackled and failed at (since the 1980s - just pick up a 1980s edition of Personal Computer to see many companies attempting the same thing). So I am now taking baby steps with a seventh attempt, by building a tool for enterprise users to build small webapps which can integrate with other systems, based on the look and feel of one of my favorite products of all time, Visual Basic 6. The product is called Pilot, and has a simple VB style editor, and allows you to build GUI apps, and microservices. It uses a simple component model based on 1 Javascript function per component, for both server side and front end components. It uses NodeJS, SQLite, and VueJS. It used to run as an Electron dekstop app but the I decided to dump Electron once I made Pilot container native so that it can run on Docker, Kubernetes, OpenShift, and Ubuntu Snap, as Electron would make the container runtimes too large. So the big focus now is to make alot of integrations so that people who work in large enterprises and just want to build a GUI really fast can build an app on top 3Scale, Mulesoft, Kong, Rest APIs, Postgres or other enterprise stuff. There have been many people involved in this project and I give my thanks to them, since they have written alot of the code. Unfortunately I am the only person who is happy to put my name to the commits as I have a very forward thinking employer, Red Hat, although there have been major contributors from Google and Microsoft as well. Also, I must stress that this project and the views of myself and other team members in no way represents the views of our employers at Red Hat, Google, or Microsoft. The project is still very rough around the edges, but please feel free to reach out to me for any questions that you may have.
- beagle3 7y ago
  Haven't had a chance to try it, but the backstory and the project both look awesome. Thanks!
  - zubairq 7y ago
    Sure thing. Just reach out to me if you need any help :)
- thinkingkong 7y ago
  I might be doing it wrong but it doesnt seem to take mobile into account at all. The website is all rendered incorrectly. Its not really a good commentary on what youve spent all your time on but first impressions are super important, as you wont have a lot of time to communicate what youre trying to accomplish.
  - zubairq 7y ago
    Yes, you are totally right, Yazz Pilot sucks on mobile right now. Our first priority is to get it working well with Google Chrome on Desktop (most Enterprise users still chained to Laptops). Second priority is to get it working on Firefox well. Third is Tablets, and last is mobile.
- pjmlp 7y ago
  I really like the idea, thanks for providing the background how it all came to be and good luck with the product.
- ihrimech 7y ago
  Very nice idea for the project. The "baby steps" you're taking seems to be the right way to go. You may need to focus a bit more on the documentation though. I was interested in trying the product (for the sake of trying), but couldn't easily go further than the install steps.
  - zubairq 7y ago
Gys 7y ago
What are comparable alternative products and why do your alpha version users think this is better?
- zubairq 7y ago
  Most people who have used this product haven't really tried the alternatives so I don't really think that Pilot is better than the others, just different in that Pilot uses Javascript as a scripting language, is easy to setup, and is Open Source. But https://www.retool.io https://www.retool.io, https://bubble.io/ https://bubble.io/, Mendix, Outsystems, Oracle Apex are all much better paid options than Pilot And https://www.lazarus-ide.org/ https://www.lazarus-ide.org/ is probably a better Open Source option Anyway, it is still very early days for this tool, so still a long way to go
- tyingq 7y ago
  I see some similarity to Google's Appmaker.
  - zubairq 7y ago
    Yeah, Google app maker is pretty cool, but I think only available as a cloud option. Maybe in the future we could offer cloud hosting of Pilot like Wordpress does as one possible revenue option
    - tyingq 7y ago
      Yes, and not only that, but apps you make in Appmaker are only visible to those that have an account in your Gsuite domain. You can't expose the app to customers or partners unless you add them as a Gsuite user. I think that was a mistake on Google's part.
      - zubairq 7y ago
        Does that mean that you can’t share appmaker apps with people using their internal company email address if their email is not hosted by google mail?
magicmouse 7y ago
I have been watching this project for a while, but now i can see the full source code. A very ambitious project as so many of the "low code" products today are basically fixed ecosystems where the components are sourced by the company, and you basically mix and match pie chart drawing modules with line graphs, etc., to make dashboards. This is more like VB6 which i understand was the goal. Many people consider VB6 to be the peak of power * ease of use.
- zubairq 7y ago
  The project has always been OpenSource, so full source code was always available. Yes you are right, I would like Yazz Pilot to be a total VB6 clone, but I still can't implement 90% of what VB6 did like Modules and inline stepping through code Aren't you the Beads guy (http://beadslang.org/ http://beadslang.org/)? I have seen it before and it looks pretty cool
dspillett 7y ago
Imitating VB6 might be the wrong direction for a lot of people here. I can see the point for devs, but not "Joe public" who has no desire to actually program. Most quick app needs for the common person trying to find better ways to do their job, in my experience, tend to be of the "I want to collect or collate some data and output some reports" and for that Access might be the better starting model, maybe taking some UX inspiration from the way people use/abuse Excel in similar contexts. Cool project though, don't let my thoughts (which could easily be quite wrong) stop you, especially if your target is people who do (or want to) dev a bit rather than people who would prefer not.
ataxexe 7y ago
I was present at two demos Zubair did and it really looks promising. I've seen some tools like that and the big non-tech problem was always how to sell. Indeed there are a lot of developers that wouldn't use such tool, but there is a huge market that can benefit from it and it's awesome to see Zubair aiming on the right spot and with one of the best approaches: Open Source! I'm closely looking forward to see Pilot reaching the cloud :)
- zubairq 7y ago
  Thanks, I know who you are dude. Actually the work you did of installing full OpenShift 3.11 (Kubernetes distro from Red Hat) on a standalone cluster which you carry in your backpack was a big inspiration to me to make Pilot portable and to work on Kubernetes (https://www.backpackcloud.com/episodes/episode-1-a-portable-cloud-provider https://www.backpackcloud.com/episodes/episode-1-a-portable-...)
rboyd 7y ago
This looks pretty interesting! Is there any auth model today or would this be best deployed behind the DMZ?
- zubairq 7y ago
  At the moment you can only protect an app made in Pilot using Keycloak or Redhat SSO. The option can be found in the top menu of the editor. In the future we already have plans to support active directory and other security models
  - rboyd 7y ago
    Perfect. I love Keycloak. Thanks!

95 points by zubairq 7y ago

Hi Zubair here, I just posted a project called Pilot that I have been working with on and off for many years now. The tool is supposed to be a Visual basic style drag and drop GUI builder for enterprise apps. the interesting part of the backstory is that this is my 7th attempt to build this! So the back story is: I first thought about building a web based development tool since I was at University where I built a Hypercard style development tool called OpenPage (using C++ on Solaris) as my Computer Science project which scored as one of the highest projects of the year at Machester University, UK (1994). My beginners confidence didn't carry on into anything concrete though, until 2000, when I wrote the second version in Java, but this didn't go anywhere. For a third version I tried to develop this in Javascript, Ruby and Sintra in 2007, but I didn't realise how difficult a problem this was. Also, in 2008 a fourth attempt was made where I tried again in Erlang, but it didn't go anywhere. So I hung up my boots and then in 2013 after building a startup NemCV with Franco Soldera I tried a fifth time and built a Clojure/Clojurescript based tool called Clojure on Coils, which was pretty cool, but most people thought it was a joke. Eventually after a couple of years I realised that I didn't understand the problem well enough, so for attempt six decided to outsource the problem by investing in a project by Chris Granger and Rob Attori, called LightTable/Eve. They tried to solve the problem of building a usable development system, but by 2018, they too realised that this was a huge task, and very hard to monetise. So, I took a step back and realised that I now had a lot of knowledge about the subject, but needed to find a way to build something that can be commercially viable as an entry point, as this is a problem that 10,000s of companies have tackled and failed at (since the 1980s - just pick up a 1980s edition of Personal Computer to see many companies attempting the same thing). So I am now taking baby steps with a seventh attempt, by building a tool for enterprise users to build small webapps which can integrate with other systems, based on the look and feel of one of my favorite products of all time, Visual Basic 6. The product is called Pilot, and has a simple VB style editor, and allows you to build GUI apps, and microservices. It uses a simple component model based on 1 Javascript function per component, for both server side and front end components. It uses NodeJS, SQLite, and VueJS. It used to run as an Electron dekstop app but the I decided to dump Electron once I made Pilot container native so that it can run on Docker, Kubernetes, OpenShift, and Ubuntu Snap, as Electron would make the container runtimes too large. So the big focus now is to make alot of integrations so that people who work in large enterprises and just want to build a GUI really fast can build an app on top 3Scale, Mulesoft, Kong, Rest APIs, Postgres or other enterprise stuff. There have been many people involved in this project and I give my thanks to them, since they have written alot of the code. Unfortunately I am the only person who is happy to put my name to the commits as I have a very forward thinking employer, Red Hat, although there have been major contributors from Google and Microsoft as well. Also, I must stress that this project and the views of myself and other team members in no way represents the views of our employers at Red Hat, Google, or Microsoft. The project is still very rough around the edges, but please feel free to reach out to me for any questions that you may have.

beagle3 7y ago
Haven't had a chance to try it, but the backstory and the project both look awesome. Thanks!
- zubairq 7y ago
  Sure thing. Just reach out to me if you need any help :)
thinkingkong 7y ago
I might be doing it wrong but it doesnt seem to take mobile into account at all. The website is all rendered incorrectly. Its not really a good commentary on what youve spent all your time on but first impressions are super important, as you wont have a lot of time to communicate what youre trying to accomplish.
- zubairq 7y ago
  Yes, you are totally right, Yazz Pilot sucks on mobile right now. Our first priority is to get it working well with Google Chrome on Desktop (most Enterprise users still chained to Laptops). Second priority is to get it working on Firefox well. Third is Tablets, and last is mobile.
pjmlp 7y ago
I really like the idea, thanks for providing the background how it all came to be and good luck with the product.
ihrimech 7y ago
Very nice idea for the project. The "baby steps" you're taking seems to be the right way to go. You may need to focus a bit more on the documentation though. I was interested in trying the product (for the sake of trying), but couldn't easily go further than the install steps.
- zubairq 7y ago

Gys 7y ago

What are comparable alternative products and why do your alpha version users think this is better?

zubairq 7y ago
Most people who have used this product haven't really tried the alternatives so I don't really think that Pilot is better than the others, just different in that Pilot uses Javascript as a scripting language, is easy to setup, and is Open Source. But https://www.retool.io https://www.retool.io, https://bubble.io/ https://bubble.io/, Mendix, Outsystems, Oracle Apex are all much better paid options than Pilot And https://www.lazarus-ide.org/ https://www.lazarus-ide.org/ is probably a better Open Source option Anyway, it is still very early days for this tool, so still a long way to go
tyingq 7y ago
I see some similarity to Google's Appmaker.
- zubairq 7y ago
  Yeah, Google app maker is pretty cool, but I think only available as a cloud option. Maybe in the future we could offer cloud hosting of Pilot like Wordpress does as one possible revenue option
  - tyingq 7y ago
    Yes, and not only that, but apps you make in Appmaker are only visible to those that have an account in your Gsuite domain. You can't expose the app to customers or partners unless you add them as a Gsuite user. I think that was a mistake on Google's part.
    - zubairq 7y ago
      Does that mean that you can’t share appmaker apps with people using their internal company email address if their email is not hosted by google mail?

magicmouse 7y ago

I have been watching this project for a while, but now i can see the full source code. A very ambitious project as so many of the "low code" products today are basically fixed ecosystems where the components are sourced by the company, and you basically mix and match pie chart drawing modules with line graphs, etc., to make dashboards. This is more like VB6 which i understand was the goal. Many people consider VB6 to be the peak of power * ease of use.

zubairq 7y ago
The project has always been OpenSource, so full source code was always available. Yes you are right, I would like Yazz Pilot to be a total VB6 clone, but I still can't implement 90% of what VB6 did like Modules and inline stepping through code Aren't you the Beads guy (http://beadslang.org/ http://beadslang.org/)? I have seen it before and it looks pretty cool

dspillett 7y ago

Imitating VB6 might be the wrong direction for a lot of people here. I can see the point for devs, but not "Joe public" who has no desire to actually program. Most quick app needs for the common person trying to find better ways to do their job, in my experience, tend to be of the "I want to collect or collate some data and output some reports" and for that Access might be the better starting model, maybe taking some UX inspiration from the way people use/abuse Excel in similar contexts. Cool project though, don't let my thoughts (which could easily be quite wrong) stop you, especially if your target is people who do (or want to) dev a bit rather than people who would prefer not.

ataxexe 7y ago

I was present at two demos Zubair did and it really looks promising. I've seen some tools like that and the big non-tech problem was always how to sell. Indeed there are a lot of developers that wouldn't use such tool, but there is a huge market that can benefit from it and it's awesome to see Zubair aiming on the right spot and with one of the best approaches: Open Source! I'm closely looking forward to see Pilot reaching the cloud :)

zubairq 7y ago
Thanks, I know who you are dude. Actually the work you did of installing full OpenShift 3.11 (Kubernetes distro from Red Hat) on a standalone cluster which you carry in your backpack was a big inspiration to me to make Pilot portable and to work on Kubernetes (https://www.backpackcloud.com/episodes/episode-1-a-portable-cloud-provider https://www.backpackcloud.com/episodes/episode-1-a-portable-...)

rboyd 7y ago

This looks pretty interesting! Is there any auth model today or would this be best deployed behind the DMZ?

zubairq 7y ago
At the moment you can only protect an app made in Pilot using Keycloak or Redhat SSO. The option can be found in the top menu of the editor. In the future we already have plans to support active directory and other security models
- rboyd 7y ago
  Perfect. I love Keycloak. Thanks!