Skip to main content

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index] [List Home]
Re: [ide-dev] Eclipse IDE popularity: Snyk report vs download stats

I just found this in their report:

"Using the IntelliJ IDEA Snyk plugin, you can ..."

Do you see any reason why Snyk should try hard to get a sample *without* a bias towards IntelliJ?

OK, they seem to have an Eclipse plug-in, too, but they don't even advertise it on https://snyk.io/product/ - only VSCode and IntelliJ.

Of course not everybody will notice this and draw corresponding conclusions.
<shrug>

Stephan

On 06.02.20 09:03, Mickael Istria wrote:
Hi all,

If you look at trends (from whichever sources), you'll see that Snyk has published its yearly "JVM trends" report recently. In this report, which is based on a ~2000 self-engaged benevolent voters, the Eclipse IDE is credited of 20% marketshare in the IDE landscape, with a decrease of almost 50% over last year. However, we have recently discussed on this mailing-list that Eclipse IDE has received a sustainable increase of downloads according to the download stats provided by Eclipse Foundation: https://www.eclipse.org/lists/ide-dev/msg01926.html , with each release receiving ~4.5 million downloads, which we estimate -a bit incorrectly but hopefully not too much- as being equivalent to the number of end-users.

So let's try to combine those together.
Facts are: Eclipse IDE download stats have shown no loss of downloads over last year, on the contrary; while Snyk reports 20%-absolute / 47%-relative loss of marketshare over last year.
Possible conclusions:
1. Both are reliable, and there have been 4.5 million new Java developers between last year and this year, and all chose to not use Eclipse IDE. 2. The Eclipse download stats are not reliable and fail at showing an actual decrease
3. The Snyk report is not reliable.

Because I trust the Eclipse Foundation, and because I do not believe the net gain of Java developer in the world can reach so high number, and because trends based on non-representative panels are never to be trusted; I obviously come to the conclusion that this Snyk report is totally erroneous, and to be more explicit, bullshit, if taken as a real market indicator.

The issue is that this report is now here and it comforts some positions and will be used by competitors (fair) to boost their momentum; and even worse, some not so rigorous architects or project managers in some companies may use this indicator as a justification to reduce investment in Eclipse IDE, as users or contributors. So this -and some similar other- report is harmful, because it contains some bullshit trends, which some influential people will not challenge and will take too seriously and use as a justification actions against the Eclipse IDE. Such influential people can typically be your boss ;)

With this message, that you can share via the link from the epp-dev mailing-list archive, I hope it gives you, dear Eclipse IDE interested parties, some counter-arguments to challenge colleagues and bosses that will instrument this report to drive to bad conclusions.

Cheers,

--
Mickael Istria
Eclipse IDE <https://www.eclipse.org/downloads/eclipse-packages/> developer, for Red Hat Developers <https://developers.redhat.com/>

_______________________________________________
ide-dev mailing list
ide-dev@xxxxxxxxxxx
To change your delivery options, retrieve your password, or unsubscribe from this list, visit
https://www.eclipse.org/mailman/listinfo/ide-dev




Back to the top