Is anyone surprised that so many of those are payed services instead of OSS projects?
I understand payroll, analytics and others like such. But things like bug trackers, dashboards, CIs, exception handling, log monitoring. At some level maybe it is a bit like selling shovels during a gold rush. Find what most start-ups need and monetize that. The gold prospector has a high chance of going home empty-handed, but he will surely need a shovel to even start digging.
Worth noting: developer tools companies don't sell primarily to startups, who represent a tiny, tiny, tiny, tiny fraction of the software developer market, frequently are averse to spending meaningful amounts of money, and have business turnover which is higher than the churn rates for many of these tools.
> Is anyone surprised that so many of those are payed services instead of OSS projects?
No. Developer time is expensive, whether you're contracting it out or you're paying for it amortized over salary/benefits/etc.
If its going to take more than 4-8 hours to setup and/or more than 1-2/hours a month to maintain, I'd rather get a credit card out and just pay for it.
Do what you're best at, outsource everything else.
I'd rather use an OSS solution but pay someone else to host and maintain it, that way if they shut down or whatever I can just copy my data and worst case run it myself.
"Standards compliant format" can be misleading. Sure, an app might give you a zip file of CSV files (which is as 'standard' and 'humanily readable' as you can get), but that doesn't mean you'll be able to import it into another tool/software. The data model could be completly different.
It does. You can run any version of any OSS programme with or without any custom patches you write. You can obviously import that zip-of-csvs into that programme.
Maybe I phrased that poorly. All data that can be retrieved or viewed can also be exported with enough effort. But OSS doesn't guarantee that process will always be easy and hosted, closed-source doesn't guarantee it will be hard.
> If its going to take more than 4-8 hours to setup and/or more than 1-2/hours a month to maintain, I'd rather get a credit card out and just pay for it.
Is it clear that paying helps you set it up and maintain it? I sincerely doubt it. That's assuming you don't need to configure/customize/extend the solution; Some OSS software are leagues easier than proprietary solutions - and have guides online and the like. Finding experienced implementers who only charge 4-8 hours to implement and 1-2 hrs a month to keep you maintained for a proprietary solution may also be a very small pool.
* Paid services advertise. OSS projects generally don't.
* Paid services usually have customer support. OSS projects generally don't.
* Paid services have nice on-ramping videos and nice installers. OSS projects have config files, install procedures, tutorials, scattered blog posts, FAQs, mail lists, etc.
* In the end, it is all about how much work you're going to do yourself. You can't do it all.
> * In the end, it is all about how much work you're going to do yourself. You can't do it all.
In my perfect I would pay for services that take care of setting up FLOSS projects and looking after the daily problems (network, storage, updates, configuration).
The code is still free and libre, the people (and their experience) well paid.
I believe this gets at the core of RMS' goal. I don't believe I've ever heard him say he's against people charging money for services, but even if he had, his point is that non-free software is inherently defective. I would also pay money for many of these services, but once I've paid I want the software to answer completely to me - don't hold it hostage.
This is a super important point. My group (the digital services division of a very large advertising agency) pays for Basecamp (pm tool) and Sifter (simple issue tracking) because the features and monthly price were better value than configuring and installing OSS tools, not to mention all the non-technical users who were already familiar with them and wouldn't require training on something they'd never seen before.
However, we use Gitlab for our DVCS (kind of surprised that's not on the list!) because that was a) very easy to install and b) better value compared to paid offerings, based on our usage. Also it's for the developers only, so ramp-up is less of an issue even though none of us had used it before.
So yeah for a lot of companies it's just about value for money. OSS is certainly free to download but the resource costs for installing/operating/training/supporting are not.
As another fan of GitLab: gosh yes, I can't recommend it enough. I also use ownCloud for my personal info management and file syncing. The only thing I don't run myself is email, which I pay Fastmail to do, because dealing with all the required things to have a decent reputation score so I don't end up in a spam box is too painful.
The only thing I don't run myself is email, which I pay Fastmail to do, because dealing with all the required things to have a decent reputation score so I don't end up in a spam box is too painful.
It's a compromise, but you could use a paid SMTP service (e.g. Sendgrid, SMTP2Go) for your outgoing mail and still use your own locally hosted system for the UI and receiving mail. At least that way, you'd have more guarantees over mail safety, etc.
I've never mass-sent emails outside of local domain, but I had set up various MTA/MDAs and from I've seen, proper A+MX+PTR records, configured SPF and DKIM, and having proper RFC-compliant emails (and, uh, obviously, not spamming anyone) should be enough to stay way below the threshold.
GitLab B.V. CEO here. Thanks for mentioning GitLab! I was pretty disappointed not to see it in the overview and a pull request from my teammate wasn't merged https://github.com/cjbarber/ToolsOfTheTrade/pull/38 It is hearthwarming to see three comments mentioning it here.
"When I first started delicious, we had to host most of the services ourselves. CVS, mail, mailing lists, etc etc etc. These days, lots of that stuff is available as SaaS. What are the tools and services people use instead of hosting their own?"
If you want to do self-hosting, there's open source available for pretty much every category. That's not what this list is for, though.
Just submitted a pull request for Collabtive (pm tool): it's both available for self-hosting (FOSS) or reasonably-priced SAAS. We'll see if the request is accepted. :)
> Is anyone surprised that so many of those are payed services instead of OSS projects?
Not surprised, as developer time is expensive. At the same tie, I find it funny and contradictory that many on HN complain about Microsoft's stuff not being OSS.
I am a Microsoft fan, but I hate when those bastards kill their products leaving people using them without real alternatives, for example: Silverlight and XNA. XNA is now re-written as an OSS _from scratch_ by very kind people (MonoGame), so even if it falls out of fashion, you could fix bugs yourself.
Actually, I don't get why analytics should be SaaS. The detailed history of all user interactions with your product since launch day is one of your company's _crown jewels_, far more than your payroll data, or even your ticketed customer support interactions. You should be warehousing and mining that event data yourself, not outsourcing the ownership and analysis to a one-size-fits-all SaaS. (Disclaimer: Snowplow co-founder :-)
Not really. I have yet to find any good OSS GUI software (don't say Firefox - I am using it but still has lots of UX quirks). Let's take bug tracker as an example. There is ErrBit and while it is in active development and fairly good it is still far from as good as SaaS-alternatives. And for high volume bug trackers, exception handling and log monitoring is something that requires a fair bit of sysop-work to handle.
It's a great list, but I don't understand why it's limited to SaaS. Self hosted apps could definitely be considered "tools of the trade". Perhaps I'm just biased because I sell a self hosted project management app - http://duetapp.com, but it seems like an arbitrary distinction....
Short story - It's a product that I needed that didn't exist. There are only a handful of self hosted project management apps and none of them felt like the right fit, especially with regard to UX. After a couple of years of waiting I finally decided to build it myself.
I wrote about some other reasons on my blog (i.e. cost), but the primary reason is that I wanted it for myself. Here are the links to the blog posts if you're interested in reading more about my thought process:
Thanks! It's always interesting to hear the thought process behind entrepreneurs choosing a path that isn't the norm (since SaaS does seem to be the default path for better or worse).
The same question could be asked for any SaaS provider. I seriously do not understand the constant circle jerk over SaaS. How many companies going tits up does it take before someone says "hmm maybe if I control it I have less risk"
Of course the same question could be asked of any SaaS provider, and I'm sure it is in the appropriate context. However, 23andwalnut did choose to go the self-hosted route so I think my question about why he chose to go the self-hosted route is completely valid and not at all a circle jerk over SaaS. You seem incredibly defensive over a simple question.
Your question implies that SaaS is the assumed/"normal"/"natural" way to go when developing software, which is basically the circle jerk I was referring to - if it isn't a format locked system your customers have 0 control over or visibility into, it isn't worth doing?
You have to pick your poison: build up subscription costs, or build up support costs to maintain your own deployments. Depending on your budget, the number of users, the need, the support contract with the vendor, and your priorities, the costs can go either way. That's the premise that both ___ as a service and ___ proprietary software selling support contracts are built upon.
How many users do you have on that JIRA instance? Guessing it's ~500, which equates to around $32 per user/y, which is actually fairly cheap considering the value add the developers would get out of it.
Fuck all extra value. 395 seats for nothing. Not only that, it's a pile of crap. It's slow, unreliable and requires so much configuration to bend it into an acceptable shape that it has cost probably 20 days' work on top of that and requires someone to spend 2-3 days a month on it. It has also gone down twice due to workflow failures and indexing bugs which have taken a day to resolve bouncing back and forth to support (they are actually quite good). We also integrate with Crucible which barely works on a good day and we've resorted to restarting it twice a day to stop it leaking memory and hanging. Plus it takes 9 whole days to index our repository after an upgrade!!!
And we can't go hosted because X isn't available and Y isn't available. Also hosted LOST a load of people's data a year or so ago.
Believe it or not, including the pricing strategy and administrative overhead we could have wrote our own tailored software for less after 3 years.
Bad product selection on our part, but despite the marketing and general buzz around Atlassian being the opposite, it stinks as a platform. Nothing but pain.
What happened to all the applications and services for outsourcing HR departments in terms of payroll, benefits, ect.? -- thought there were a bunch more than Workday and ZenPayroll?
Getting a full list of those would be particularly useful.
We tried to use Airbrake at the last startup I contracted for and it was embarrassingly underwhelming. The landing page makes it look awesome, but compared to HoneyBadger, it was just not up to par when it came to error notification. We were looking for centralized errors for Ruby and Python code so we ended up going with Rollbar.
Nice job Chris! If anyone wants to see feedback from other developers about these tools, the majority of them are on http://leanstack.io/popular (disclaimer: Founder of Leanstack.io here).
Nice app. I noticed vendors can't self-select a category. That might be a useful feature for services that want to identify themselves in a category that isn't immediately clear. For example, the lines could blur between File Storage and Cloud Storage.
Thanks! Self-selecting for categories could get tricky because we're regularly creating and deleting categories as we see fit. We'd rather stay on top of that than put the burden on vendors. But we usually do make edits if a vendor asks.
I think it's because the list is being built by pull requests. Two people submit a Log Monitoring section at the same time and the repo owner merges without really paying attention.
I created a PR for a Cloud Storage section since I didn't notice one yet and added https://kloudless.com to it. (disclosure: co-founder of Kloudless here).
Google Docs is probably the closest cloud storage-related utility on that list, but is under the "Notes" category.
It's fascinating that the proliferation of EC2 has enabled services by different vendors to be provided on virtually the same physical network (with associated performance).
Much of what is attributed to the rise of SaaS and "The Cloud" is in fact a direct consequence of Amazon's infrastructure.
As a small time developer, none of these tools is in my budget, I have no budget. So this list is useless to me.
Ok sure, I wasn't your target audience, that's fine. but the title doesn't suggest "Tools for active/established companies who have money to burn and needs to fill"
You guys seem to be missing Wercker. http://wercker.com/ is my favorite CI/CD. Its free while its in beta, has a great UI and has a simple but powerful configuration system
I missing the web directories of yore that have such nice lists. Search engines only bring up a few relevant results and the rest are all keyword and link stuffed. Hope you take some pull requests or make this a wiki.
This is a very valid point, the other day I was looking for a wifi sniffer tool because I was getting terrible speeds over the last couple weeks and I wanted to ascertain the cause; there was nothing recently put out that compared more than 2 products, nearly all the blog posts were from 3-4 years ago, and the tools they compared were either bought out by cisco and the like or went out of business/ended support. really just having this type of list would be huge.
I wish there more open-source libs offering commercial options. I don't want to see libraries I depend on being abandoned. Might also want to have access to support. I'm sure there are many companies who would pay and still want the project to be open source and have an active community. Sidekiq is one example of this which seem to be going quite well. I think it is a problem that very important projects are being developed for fun after work (openssl anyone). While I do not want the "pay for all libs" like in the .net world I do hope to see more for-profit open source projects.
Not everyone is a startup working with node or RoR. Most companies with more than a handful of staff would rather pay someone else to make sure it doesn't break.
Issue tracking? I have seen a private BitBucket repo with a public list of issues. BitBucket provides private repos for free, unlike Github, by the way.
I understand payroll, analytics and others like such. But things like bug trackers, dashboards, CIs, exception handling, log monitoring. At some level maybe it is a bit like selling shovels during a gold rush. Find what most start-ups need and monetize that. The gold prospector has a high chance of going home empty-handed, but he will surely need a shovel to even start digging.