Show HN: Self-hosted project management application | Hacker News Reader

Show HN: Self-hosted project management application

92fiveapp.com

135 pointschintanb81 comments

Free 5 minute grumpy feedback on the landing page :)

Requires Javascript to display anything for no good reason. After enabling it and reloading it takes more than 10 seconds to anything but the loading icon to appear. When I quickly scroll down to see what's there, there is nothing, then it fades in. Screenshots are tiny. When I click them they jump into my face. FAQ is hard to impossible to read with the color combination and font and centering. On the whole page I have no idea what is a link and what is not. FAQ items take me back to the top when I click them. Requiring a mail address for downloading (free?!) software is a no-no for me. The buttons at the bottom (fb, twitter etc) have no anchors set so if I hover them, I have no idea what clicking will do. I would not want them to open my mail client or something. No idea what the rightmost icon even means.

> 92five app by Chintan Banugaria is licensed under a Creative Commons Attribution-NoDerivatives 4.0 International License.

Creative Commons are not for programs. Please choose AGPL or something like that for code.

You waste a lot of vertical space, I would suggest putting elements of the top-most part of the side closer together and maybe strip some text like this: http://i.imgur.com/9MjNLvy.png That way the full screenshot has a better chance to appear on the page.

Writing:

Don't use slashes when you can use a word instead. "No-one can see / access your todos." could be "No-one can see or access your todos.". You are using "to-do" elsewhere which I prefer. I would recommend also using "To-Dos" in that header.

You say "I am sure you will love the design.". I would not use "I" in that page unless you introduce yourself first.

Also spotted "Yes its free." -> "Yes, it's free."!

-----

The praise is hidden here! The product itself looks slick and useful. Personally I am not a fan of flat design but you seem to have pulled it off nicely. Self-hosted tools are the best, thanks for doing that! I really really really suggest you make the landing page less annoying though. :P

I agree with the JS critique. I actually thought something was wrong when it didn't load. I'm not sure the loading time is worth it for those pop-in effects as they aren't really necessary. Also I'm 28 with pretty good vision but I'm having a hard time with that lime green on white.

Yeah I found that loading indicator bizarre. What am I waiting for? Is it circumventing the browser's rendering engine and doing it all in canvas or something? (I mean, I know it's not, that would be silly. It's more likely having to make multiple network requests to fetch the rest of the content.)

I can understand the other javascript-ey bits as forgoing some amount of UX/usability in exchange for flashier visuals. But the loading indicator on initial page load just doesn't make sense to me.

The intent of the license (no derivatives) seems very different from AGPL (strong copyleft).

Please don't recommend AGPL uncritically. Not everyone cares to think through the implications (yes, I have had that discussion with someone who told me was fine about me using his AGPL licensed program in a clearly non-AGPL compliant way because it was just GPL3 but for the web. I Had to find another solution.)

I think this app has potential but after installing a local copy and playing with it for a few minutes there are a few things right off the bat that are a bit of an issue:

First, creating a project doesn't work, after putting in the required info and clicking create I'm taken to an error page with no feedback as to what actually happened, even though the screen says 'something went wrong and we've noted that'. Also, requiring to add collaborators even if it's yourself is redundant, this shouldn't be required.

Tasks - The main tasks screen really isn't useful at all since each 'Task' takes up a huge amount of space, and all the links on the task card don't do anything, and even when sub-tasks are entered they don't show on the card. This section really seems like it should be tied into projects, not a standalone section, especially since there's a 'Todos' section as well.

In many places where input is required, it's not immediately apparent that the colored title area is editable or requires input.

There's quite a few little UI tweaks that need to be made, for instance the line height in the quick notes section on the dashboard does not match the notebook lines, so typing anything in there looks sloppy.

Took 15 seconds to load the landing page. Honestly, you lost me as a user after the first 5 but I wanted to see the rest of the page.

Scrolled to the bottom quickly, had to wait for it to animate in so I was staring at a blank page for a few seconds.

Can't be bothered to try the product, since I'll be old and gray by the time it boots.

It's probably loading slowly because being linked on HN is giving the server too much traffic. It's self hosted anyway, so that doesn't really matter.

It does really matter - the landing page loads 1.9MB of content. This is easily 10x the size one should expect from a landing page. It's reasonable to assume that the app itself would follow similar proportions, and that means (most likely) load times upwards of 45 seconds. No thanks.

45 seconds? That sounds like a huge overestimate, especially after you've cached the assets. A reasonable internet connection should shift 1.9MB in 2-3 seconds.

Depends which country's reasonable you are talking about. In my country, broadband is anything more than 256 kbps! So, to call their services broadband, almost all providers drop speed after FUP limit (anywhere from 8-200 GB) to 256kbps.

And average speed is something like 2-3 Mbps.

Shitty internet is your problem. Fix it so you can experience the internet properly in 2014.

Or devs could stop including fucking huge but pointless assests in what should be a simple web page.

"Test on real hardware" includes testing on the reduced bandwidth that most people have.

Don't forget to make sure it works on IE6 and on an ISDN too.

The only problem here is shitty internet in 2014. No civilized country has any excuse. Running fiber is incredibly inexpensive relative to almost all major infrastructure costs. A 2MB web page is not the problem.

Don't forget to make sure it only works on 48 core supercomputers equipped with 8 TB of RAM.

The only problem here (and everywhere) is shitty developers thinking that being in 2014 is an acceptable excuse for extremely shoddy work. No self-respecting developer has any excuse. At the global scales, it must be costing trillions of dollars in wasted electricity. All 2MB web pages in the world only compound the problem when they should be < 100kb.

As with everything, there should be balance. This certainly isn't it.

Damn shitty developers like John Carmack wasting computing power on better graphics! Where's my pong!?

I would disagree. Why are pages so massive? Is the page displaying pictures and text? Like it did in 1991?

If so, why is there so much JavaScript and other fluff loaded?

The increase in average pages sizes isn't really necessary. Everyone appears to be jumping on the "let's load LOADS of data" bandwagon, which isn't much fun on a mobile connection, let alone a 2G connection.

And before you say that 2G is unacceptable, let's see how well most rural areas cope, or how well you do in a valley in Wales.

Yes, poor Internet connectivity is not great, but let's not foolishly assume that everyone is on a fibre connection. I am, but others are not, and I certainly wouldn't write software in the belief that they are running a beefy development machine that I have, nor would I write software in the belief that everyone has the same network connection. (Would the argument be "buy a better computer" really be acceptable if software that I wrote ran slowly on someone's machine due to stupid non-optimised loops????)

Not everyone is experiencing the Internet the same as you, so don't assume that they are.

If that's the case, why do we make thumbnails of pictures instead of just using an img width tag and resizing giant pictures down to thumbnail size?

1.9MB is as large as some ENTIRE pieces of software executables (though they may be dynamically linked).

EDIT: Not meant nastily, more food for thought.

It took 49 seconds for me to load the __landing page__.

It would have been done in 3 seconds (which is still an insanely long amount of time) except that this image: https://i.creativecommons.org/l/by-nd/4.0/88x31.png took 49 seconds to load.

Chrome's network panel shows:

    Blocking   27.15 s
    DNS Lookup 0.137 ms
    Connecting 21.76 s
    SSL        20.59 s
    Sending    0.129 ms
    Waiting    78.344 ms
    Receiving  1.252 ms

For a damn 5kb image.

Loaded fine for me. Rest of site looks good.

Will definitely try this out. May sound strange to others but I could use a project management software, even though I'm a one man project operation...

Okay, I was just reading through the landing page. Here are my thoughts:

* I'm not sure that 'self hosted' needs to be mentioned in the heading. You don't mention cloud or SaaS anywhere and have a big 'download' button so it should be obvious without stating it.

* "92five app covers all the basic project management features". Don't say your product is basic, the features should rock.

* Responsive. Yawn. Most managers don't know what this means. You don't mention 'mobile' anywhere in this section.

* Built with. This is the stuff you're interested in, and it's important but it has nothing to do with the product. Most users wouldn't care whether it's PHP, Java or Basic.

* Screenshots. Great we are past the techie bits and back to the actual product... But I don't know what I'm looking at.

* Wow, it's free! That's the best part and you mentioned it last... and in the FAQ. This should be up the top of the page!

I hope this helps.

> Wow, it's free

It is not free - you can't make derivatives. It is just free of cost.

I'm really impressed from what I see. Would love an excuse to not spend $200/month on project management software :)

I use a very popular SaaS right now but would actually rather self-host (I like to own my data.)

Haven't had a chance to install it yet but from from the look of it and general feedback I could see my company making a switch if it all checks out.

I don't need responsiveness, HTML 5, a mobile app or any of that. Just gimme something that works!

What SaaS do you currently use for project management?

I didn't want to say here, but Basecamp. And I have absolutely nothing against them but I wouldn't mind saving a grand or two every year if this can do the same thing and hold up to a dozen or so users.

I'd love that as well. But the killer feature that Basecamp has for me is that you can reply to their emails. Half of my clients have created initial accounts and haven't logged back after that. Yet, they can communicate. If this can do that, I'm switching now!

You see I didn't even know about that feature, definitely something I 'd like to look into using.

Is "self hosted" considered a positive or negative these days? It seems like there is a small niche that still wants control over their data but everyone else seems ok with the cloud.

I consider it a plus.

If I were developing a product like this, I'd build it for self hosting and charge a nominal cost plus yearly a maintenance/support subscription. Then you could deploy the same product yourself and offer it "as a service" for a monthly subscription. Atlassian does this for many of their products.

I skipped over the free part too. Makes sense to be free and self hosted but this would also be cool as a "deploy to heroku" (or whatever cloud PaaS provider officially supports php). I still wonder if the creator has plans to monetize it?

Also as for the business model you've mentioned. I've seen this done in competitive spaces like affiliate marketing. If you don't know most affiliates are super paranoid about thier data as leaking info about a campaign can lead to tons of copies and there goes your revenue. They seem to mostly prefer self hosted apps. Imobitrax offers a self hosted paid tracking system with a monthly fee. A lot of people used to use the free prosper202 self hosted tracking system too. Now recently I'm seeing a shift where people are not afraid to use cloud systems like voluum. I'm not sure what is causing the acceptance to change. It could just be that everything is cloud hosted now so people are getting used to the idea.

The problem I see with paid self hosted PHP apps, in this case, is that they are often using ioncube obfuscation and then if I can't read the source how do I know my data is more secure than a cloud provider?

I built a similar project management app (http://duetapp.com) and people love the fact that it's self hosted. This segment of the market is definitely smaller than the segment that prefers SaaS, but it's bigger than you might imagine.

Hey! I have always one doubt about this type of product (pay once). How do you distribute your source font, as well the installer to your customer? Do you use another software for a restricted area to download? Or is the download open to anyone, but you have to offer a license key to run the app?

The app is distributed as a single zip, but the installer requires a license key. Updates are distributed through the duetapp website, which also requires a license key before the update is delivered.

Nice! But is it secure? As the customer already have the source code, what stops him to change the license code? I mean, if the source code is available, I can install it by myself without using the installer.

I know a lot of PHP projects that do this use IONcube loader and mcrypt. So you get the source but it's encrypted.

Nice app. Just a question about "Free updates for a year", what's the cost after a year?

It's definitely a positive, especially for enterprise. The cloud for enterprise means your' depending on people who dont know to do a good job, and it means you spend 3x as much time communicating vs getting work done.

I don't know. I work for an enterprise company and we're coming around to the idea of putting things in the cloud. I actually think the people running cloud hosted software are more of an expert in the software and devops than the people inside your organization in a lot of cases. Plus large enterprises have so much bloat that it can be way more expensive and time consuming to launch something internally. The biggest issue seems to be that these companies are still afraid of who controls their data.

Sometimes. All I can speak about is my personal experiences with a handful of enterprise customers and the times we moved things to the cloud. Communication is always an issue, we find that instead of spending time in house developing, deploying, administering, etc... we spend that same amount of time either one the phone or writing emails and documentation on what we need or what is wrong or what we need done.

We also find that the what the cloud team finds as a sufficient level of quality or service often does not match what we require or expect. And fixing that is a major PITA because it comes down to culture, and suddenly you are not dealing with your company's culture but a 'cloud' culture.

And on top of all that there is the issue of who controls your data. Is it secure? is it safe? What's your uptime, downtime, maintenance? Often we get promised good things (sales people are good at what they do), but it rarely works out as such.

The answer to this is find another cloud provider, but I dont want to spend months RFI, RFP, migration, integration, etc....again because the first cloud people were terrible. It's often just better in the long term to do it yourself.

It's a pay me now or pay me later scenario and I'll live by the words if you want something done right, do it yourself, especially for enterprise.

Self hosted software generally tends to play better with others in my experience.

SaaS apps seem to be about lock in wherever possible with just the barest lip service given to standards and true interoperability.

>these companies are still afraid of who controls their data

You make this sound somehow irrational.

For a large number of companies, business data is the key to the running of the business. Why would you ever let someone else control this data, much less someone who themselves doesn't have full control?

I wasn't trying to say it's irrational. I realize why companies would be afraid of putting proprietary data into someone else's hands. I still think the idea that it's going to be stolen or leaked from a cloud provide is not likely to happen. I'd be willing bet most of the time IP is stolen it's someone inside the company doing it. So much easier for a disgruntle worker to do it.

Who says it's about theft?

Do you not remember what happened with Codespaces?

How many times are there posts to HN: "AWS is down in <FOO>" with dozens of comments "yep, XYZ doesn't work right now"

How many service based companies have been bought out by a bigger fish only for the service to be shut down or changed in some dramatic way?

For the majority of businesses I very much doubt that theft is the primary concern with using SaaS. Control, visibility, access to your own data regardless of the app/system author's current situation, etc.

Even if theft is your primary concern - and I will grant you that for most small companies where it's targeted theft, the finger more likely points to an insider - using SaaS, especially if it's one builds on a third party provider's stack, means you are increasing the potential for breaches, and you don't even know what that increased potential is, and potentially you may not even know if a breach occurred.

Imagine a company that runs a service on Heroku (who in turn have built on AWS). So by using this service, you suddenly have great potential for individuals at three other corporations to heavily influence your ability to do business and remain profitable, either through data loss, downtime, or the aforementioned security breach.

I haven't even mentioned the issue of data privacy here - it seems quite common for modern SaaS startups to rely on VC funding to offset the costs of providing their service, with the vague notion of "if you build it, they will come... and then we can somehow turn them/the information we hold on them into a profit"

While some people will seemingly climb over their own mother in their rush to hand over their personal and business data to every man and his dog, some of us still have our sanity.

Having it available as self hosted doesn't exclude the possibility that the author or another company will offer it as a service. The reverse is often not true.

The terminology is perhaps a little messy, but I think "self hosted" and "in the cloud" are not necessarily mutually exclusive. On one hand, "in the cloud" can mean a SaaS platform where your data is locked in and you have no control over the application. However, I could also own the code and data and host an application myself "in the cloud."

This is the ideal compromise to me. I can own my data and the application, but I don't necessarily have to deal with all the annoying details of managing my own servers.

Yeah when I said "cloud" I really meant SaaS. You can certainly deploy a self hosted app to a "cloud" PaaS and get a greater ability to control your data than SaaS while still getting many of the other benefits a SaaS service offers.

I'm working on a self-hosted product.

My market research suggests that the market for hosted products is much larger but the market for self-hosted products these days is underserved.

There are people who for various reasons refuse to use any hosted product (legal issue, boss said no, etc.) and depending on the competitive landscape they may not have any other options than buying your product. This produces a captive audience.

Offer self-hosting for a single fee (or for free), and offer hosted for a monthly price. That seems to be the trend.

I would really love to see some good self-hosted Basecamp killer. This is getting close, but lacks Basecamp's most important feature (at least for me): discussions. This is something that we are using A LOT, and no other software even comes close to what Basecamp has to offer in that area.

I have setup a demo env for those interested:

http://92fivedemo.pilgrimbreak.com/

Email: demoguest@92fivedemo.com

Password: 12346#0azeRtu

I'll keep it up for 1 or 2 days.

I can't seem to be able to access your demo: "This webpage is not available" .

I had no problem connecting. Did you disable javascript?

i cant access it as well ... but i believe its that my work is blocking certain ports. That could be a reason ...

The server at 92fivedemo.pilgrimbreak.com can't be found because the DNS look-up failed. DNS is the network service that translates a website's name to its Internet address. This error is most often caused by having no connection to the Internet or a misconfigured network. It can also be caused by an unresponsive DNS server or a firewall preventing Google Chrome from accessing the network.

It seems the DNS slave was not up to date so you have 50% chance of having no DNS resolution. It appears to be resolved now as I can curl the page from my home server.

It seems some people are having difficulties accessing the subdomain, which I just created. It works from work for me, but not from my home server.

I'll try to sort this out ASAP.

it seems someone changed the password...

I just logged in so password is working fine right now.

Working now, thanks.

Yeah. I just screwed it up. Apparently you can suspend your own user account.

Which is an incredibly useful option.

Very helpful thanks!

Seems like a pretty intuitive interface with decent metrics. Pretty close to the sweet spot between too simple and too hard.

I just deactivated the demo. Hope everyone had time to test the app!

Im currently working on something similar, but focused on the tasks more.

Its like a mix of a project manager and a timesheets portal. The front page contains your list of active tasks for the day, and any sticky tasks that you have. The team leader can assign tasks from projects to individual users, and be able to see the current project assigned time, and set threshold for hours spent / left.

Project managers will also be able to assign projects to teams, and also see current progress of tasks completed, time spent on tasks etc.

I was also thinking of distributing it as a self hosted open source app and maybe a hosted solution as well. For something like this, what would the recomended license be? Something like AGPL or BSD? How would any of you release the software, and if I would like to offer a hosted service in the future, would you worry of others using your software and selling the service as well?

AGPL feels toxic to me. At least add a commercial option then.

I'm working on something myself and my contributions to open source has so far been in sending and getting accepted a few pull requests with improved docs on one project and improvements to some javascript library as well as sneding a few bucks to a couple of projects that takes donations.

Haven't release anything serious yet but I know if I do it will be permissive because at the moment I'm giving it away I want it to be as useful as possible and not tied up. (Unless I want to run some dual licensing scheme. ; )

Interesting project!

I'm currently working on a green field project with a very small team on my free time, so I installed 92five on my server and will be testing it "as if in production".

I will add any issues I encounter on their Github - https://github.com/chintanbanugaria/92five

Seems like it's getting overloaded right now as I just get a loading animation for 20+ seconds.

I will take a look at this later though.

I have been liking scrumdo which also allows you to self-host the application (in addition to offering to host it for you - which is their profit model)

I just created a demo installation of 92fiveapp at Terminal.com. You just need to spin up a new container using this snapshot: https://terminal.com/tiny/yTUnxPVAiz

That was helpful. Even though the site required login, but well appreciated.

Looks great, has potential: nice design, self hosted, responsive UI, open source. Maybe create a companion offline app for smartphones/tablets that would sync with the server?

This looks exactly like something I've been looking for!

Curious, what kind of database (if any) backend is it using? If it's not using MySQL or anything heavy like that, I'm sold!

From the GitHub readme:

Minimum Server Requirements

PHP 5.4 or greater

PDO PHP extendsion

MCrypt PHP extendsion

GD PHP library

MySql Database

If your webserver is running Apache then mod_rewrite should be installed

Let me know what kind of database you want to use ? I might be able to help you out.

sqlite would be great, or some NoSQL solution such as Mongo (though I see as that being less likely).

How is mongodb "light" compared to mysql/percona/compatible ?

It's not. I just have an instance of MongoDB running on a separate server.

Guys, 92fiveapp is on github Visit https://github.com/chintanbanugaria/92five

I'd love to see a live demo of this...

let me know your email and would set up a demo account for you to test.

Where can I email you at? I'd love to demo this as well.

The app looks great, but how are you going to make money from this so you can sustain it?

This looks great.

Whoooa and it even runs on a macbook. sorry but I can't help but feeling screenshots inside "macs" look cheesy.

Reply on news.ycombinator.com