JetBrains Omea

» see jetbrains.com/omea as it lived, on the wayback machine

~ the obituary ~

JetBrains sells tools to people who write code, and has done since 2000. Omea is the exception, the one product the company aimed at everybody else, and it is buried here because the experiment ran for four years and then stopped. There is no scandal in it and no acquisition. A developer tools company built an information manager for office workers, found out that office workers do not buy information managers, and gave the source code away.

The free half arrived first. Omea Reader shipped on 4 October 2004 as a Windows feed reader that also took newsgroups and web bookmarks, and it was good enough that people used it for years. Omea Pro followed on 15 December 2004 as the paid product, and Pro is where the actual idea lived: it pulled your email out of Outlook and your conversations out of ICQ and Miranda, along with the contacts and the Office documents sitting on your own disk, then indexed all of it into one searchable pile with a tag system over the top.

The pitch was that your information has no natural home. A thread about a project exists partly in mail and partly in a file somebody sent you in March, and no single application can see across that boundary. Omea could. You could put the same tag on a mail message and on a PDF, then pull both up in one list, which in 2004 was a thing almost no consumer software did.

It was built on.NET when.NET was young and ran on Windows only. Development stopped at version 2.2. Around January 2008 the users worked out that nothing further was coming, and the complaints on the blogs of the time are less about the software than about being left holding a database in a format nobody was maintaining. On 13 May 2008 JetBrains put the whole thing under the GPL v2, which is the date this grave carries, because that is the day it stopped being a product.

I knew the company and not the product, which is the honest position on this one. By the mid 2000s JetBrains meant one thing in the circles I was in, and that thing was the IDE, an object with a very specific reputation among people writing Java. That the same company had a second life selling a personal information manager to accountants is something I learned long afterward, while looking for what else they had shipped. By the time I had a broadband line at home and a reason to read a lot of feeds, the reader was a page in a browser and not a program you installed, so the window Omea wanted to be was already taken.

~ the case file ~

~ everything in one window ~

The architecture is worth describing because the industry keeps rebuilding it. Omea ran a local index over anything it could reach, and the plugins that fed it were the whole design: one for Outlook mail, one for RSS and Atom, and others down to the file system itself. Everything that came in became the same kind of object, with the same tagging and the same search across it, whatever it had been before it arrived.

That is the shape every desktop search product of the era was reaching for, and most of them lost. Google Desktop shipped the same month as Omea Reader and was dead by 2011. Microsoft spent years on the file system that was meant to make this native and shipped none of it. The pieces that survived went into mail clients and into the operating system search boxes, so the function is still there and the category is not, which is the usual fate of software that tried to be the layer above your other software.

~ the code came back ~

The open sourcing in 2008 was not much of a handover. The repository took five commits between March and August of that year and only three of them meant anything, which is the signature of code being written somewhere private and dropped over the wall in batches. No community formed around it. Version 2.3 sat there as an early access build and nothing moved for sixteen years.

Then in 2024 a developer working under the name F. von Never pulled the 2008 repository onto GitHub and made it build again. That took a Windows XP virtual machine and a copy of Visual Studio 2008 just to confirm the original still ran, then a port up to a modern toolchain, including translating pieces written in Managed C++ and in JScript.NET, two dialects Microsoft abandoned long ago. It compiles and reads feeds on current Windows. The product died in 2008 and the source is now in better repair than it was on the day it was released, which is a fate very few things on this site get.

@}-->--

              )
             (_)
            .-'-.
            |   |
            |   |
            |   |
            |   |
          __|   |__   .-.
       .-'  |   |  `-:   :
      :     `---'     :-'
       `-._       _.-'
           '""""""

« previous  ·  random  ·  next »