Larry Osterman's WebLog

Confessions of an Old Fogey
Blog - Title

Why does Microsoft "Time Bomb" its beta releases?

Why does Microsoft "Time Bomb" its beta releases?

  • Comments 24

One question that periodically comes up is "Is <x> beta time bombed"?

First off, what's a time bomb?  It's a chunk of code that's intended to disable a beta release sometime after the beta ships.

I believe that all MS beta products have to be time bombed, I know that all the products I've worked on recently have been time bombed.  The time bomb can be mild (you lose the ability to send or receive new email), or it can be severe (the product refuses to start/boot), but all beta products are time bombed.

As far as I know, the reason is buried deep in the roots of Exchange.  When Exchange 4.0 shipped (back in 1994ish) there had been several previous beta releases.  What Microsoft didn't realize at the time was that these early beta releases of Exchange were "good enough" for the sites that were running that beta.

But the beta had a bunch of bugs in it - not serious enough to make the product unusable, but enough to cause interoperability problems with existing email systems (I really don't remember the details, but the problem was something minor like that the SMTP gateway generating uuencoded TNEF blobs instead of converting to MIME, or something like that).

We fixed the problem long before RTM, it only existed in the one beta release of the product.  However these beta servers continued to be run by companies that had received the beta for several years after we released the product.  The consequence of this was that Microsoft continued to have people reporting that Microsoft Exchange was producing illegal messages.

The thing is that no shipping product had those bugs.  The problems were seen because some very small subset of customers hadn't upgraded to the shipping version of the product.

As a result of this (and other similar problems), Microsoft started time bombing beta releases - that way they Microsoft can guarantee that beta releases don't cause problems long after the product RTMs.

  • I'm really surprised that some people are dumb enough to use beta software on production servers :| !
  • Plop said:
    >> I'm really surprised that some people are dumb enough to use beta software on production servers :| !

    Hey, there are people dumb enough to use open source on production servers - stuff developed by _volunteers_ in third world countries and Europe, taking revenue streams from genuine American companies. For the good of our nation, I think Microsoft should lobby hard to outlaw open source on constitutionality grounds (after Eldred has ruled that copyright exists for a profit motive instead of the liberal left 'good of society' whining). They should also patent everything in sight however obvious so their pathetic and insiduous communist ideology is extinguished once and for all.
  • Red Blooded American said:
    >>Hey, there are people dumb enough to use open source on production servers

    And I'm rather amused that someone can be so narrow minded like you! From your speach and name I must assume you're American... And a dumb one, indeed, start looking the world around you!!

    First, third world countries are not places filled with stupid people, there are lots of good scientific works done on them, in particular, India is known for providing the computing industry with a lot of talented programmers (note, I'm not indian).

    Second, Linux, perhaps the most known open source software, was created by a Finnish man, as far as I can tell, Finland is NOT a third world country.

    Third, there are lots of AMERICAN companies that make loads of money from open source software, like Red Hat, IBM, Sun and Novell, they support several projects, are they mad or dumb? I don't think so...

    Forth, did you know that Apache is the most used web server in the world? Not only it's the most used, it's also the less exploited, oh, wait, that's another "third world project run by volunteers"!

    Err, again, I don't think so, the project is run by the Apache Foundation, from their FAQ:

    " Is the Apache Software Foundation (ASF) a Corporation?

    Yes, the ASF is a membership-based corporation registered in Delaware, United States. It is intended to be a registered non-profit charity, and in fact was given 501(c)(3) status by the U.S. Internal Revenue Service."

    Did you see? It's a US based corporation, not a bunch of third world nerds as you implied and they gave us a much better web server than MS did with their bottomless pockets!

    On other examples, the KDE project was born in Germany, Gnome is a US product, several Linux distros were created in first world countries (Suse was german before Novell bought them, hey, Novell is american!!! Mandrake is french, Red Hat is also american, Debian is a community project based in the US).

    Of all major Linux distros, only Conectiva was from a third world country, namely, Brazil (yes I'AM BRAZILLIAN, and I think your comment is very offensive, not only to third world people but to everyone from first world countries as well that contribute to several open source projects).

    And some months ago, Mandrake acquired Conectiva and the new distro is now called Mandriva. I acknowledge that Conectiva was not very known outside Latin American, from what I've read, but their staff are very talented and technically capable, that being one of the reasons Mandrake bought them!

    That said, I can understand that as an american you're mainly interested in the good of your nation, as I'm interested in the good of mine!

    But do I think you're really narrow minded if you think that only Microsoft can do good for America!! As said before, IBM, Novell, Sun and Red Hat are also american companies and I'm sure they would fight to death against such a thing as outlawing open source software. Remember that the Free Software Foundation is ALSO american and they wouldn't let this go without a hard fight!

    As to patenting everything, US law IS stupid in these regards, so "double clicking" is something highly innovative and should be patented! Imagine how the computing world would be if, for example, MS did such a thing, everyone else would have to pay them for using this design or else they'd had to invent something new and completely different, or stick with the old world of command line interfaces! SOFTWARE PATENTS HARM INNOVATION!!!! They only do good to the patent owner, in your suggestion MS, and they would do no good to american PEOPLE in general, only to MS shareholders!!!

    Hey, I bet you're a MS employee and shareholder!!!
  • I think "Red Blooded American's" comments were meant to be sarcastic...

    And I'm not really suprised to see people using beta software on production servers. You'd be amazed some of the things people do, sometimes :) I think it's perfectly reasonable for Microsoft to put time bombs in all it's beta software.
  • I have participated actively or just surfed passively a huge amount of chat/blogs/forums. Of all stupid ignorant self-righteous people I have encountered you really, really, really are the worse.
    What makes you think that US has any kind of "right" to monopolize software developpement?
    May I presume and partake to your education with the following bit of information: the world wide web, HTML markup language, the first webserver and the HTTP protocol are ALL OF THEM "stuff developed by _volunteers_ in third world countries and Europe, taking revenue streams from genuine American companies", in other words european products, invented by a european name of Tim-Berners Lee, working for an european enterprise (CERN). Mr Lee wasn't even a computer specialist, his area of expertize was fysics.

    But you probably just confused this site with one of you KKK meetings. You guys still bravely hiding under oversized bedsheets?

    C.
  • When starting to read Antonio's comment, I was thinking, "I thought Europeans were supposed to understand irony." But then Antonio said he's not European, he's Brazilian. That explains it. Same hemisphere as the US. No wonder he didn't understand the irony.

    By the way, sure betas need to be time bombed. But it sure would be nice if source code, configuration files, and other stuff that were developed during the beta to help find what needs fixing, would be migratable to the release version instead of requiring rewriting from scratch. (Though of course there are higher priority situations that need exactly this kind of fix, like making VB6 programs migratable to .Net instead of rewriting from scratch. Even if existing VB6 programs could be migrated automatically to C# instead of to VB-- and instead of being rewritten from scratch, it would be a big help.)
  • My opinion is that people who expect other people from all over the world with different backgrounds and in different situations to understand irony or subtle jokes and trolls... Well I guess they do not (expect), they are just sad trolls trying to get a few chuckles.

    Replace the ... with your favorite insult.
  • Norman, is that the left of the right hemisphere? ;)
  • Surely the real reason is that, if customers use the "free" beta copies, they're not paying for the RTM product. And that means loss of revenue. Or am I being unduly cyncial?
  • Antonio, I don't think Novell and Sun are good examples of successful corporations. Sun... is doomed, Novell is getting there...
  • Antonio, you've been trolled. Nice one, RBA. :-)
  • To make this post and claim the reason for time bombs is purely a technical quality reason, and to not even mention the fact that someone using a free beta isn't paying Microsoft hundreds of dollars, is to ignore a huge part of the reason for time bombs.

    Has Microsoft's PR department invaded Larry's office? Let's hope not. Pretending the time bombs are there only because you care about the customer is a line that most of us won't fall for.
  • Seems I never can keep it short...

    While I understand time-bombs are in both beta's and trial versions, I think I also understand at least a few of the reasons some might continue to use older betas without time-bombs.

    For something as Exchange, I'd expect the company having participated in a/the beta program to have enough money to spare, and likely even did pay for "the real deal" once it was released - even if they didn't install it.

    However, to truly be able to test a beta (the very reason to use it - to find bugs) you sometimes have to put it in production servers.

    As a responsible sysadmin you don't want to mess with stuff if it works (especially MS software that comes with so many undocumented dependencies and so little, if any, documentation it can break if you just sneeze). Exchange has a history of doing very nasty stuff to your data (incl. if you try to update), why not updating a beta without a timebomb seemingly doing its job could be seen as doing The Right Thing(tm).

    Adam Young,
    No, I don't think it can be seen as loss of revenue (for Microsoft) if you have done what's expected of you as a beta tester. OK, I'm thinking of a smaller scale here. I wouldn't expect a company to install a beta on a farm of thousands of servers. :-) If they did, then yeah, that could be a revenue loss for the licensor (i.e. Microsoft).

    To explain why I think it can't be seen as a loss:
    I don't know how MS works today, but from my participation in MS betas I've only experienced generosity (!) if they get good and precise bug reports (again, the very reason to run a beta). I got "burned" CD-R's sent across the pond by overnight flights and hand-delivered to my doorstep. I got copies of not only the software I've been testing, but also seemingly unrelated but still useful software as a "thank you". Once a/the final build was completed, I had the RTM way before it turned up on MSDN (this was before the MSDN d/l was viable).

    With that in mind, a user or company being *active* in beta-testing a particular piece of MS software and happens to run it on 10 production machines (to get it properly stress-tested) should IMNSHO also get at least 10 licenses for their efforts. What could be missing in the equation is that MS should then also spare no effort explaining _in detail_ to such beta-program-participants what is changed and updated from the beta version tested and RTM, how and why an update to RTM could fail, and provide (for a short duration of time) a "hot-line" should the "upgrade" from beta to RTM mess up the system (here I'm thinking again of programs such as Exchange or MS SQL Server where loads of data could be at risk).
  • There is also another facet to this. It is extremely rare for a company to make a new version of a product that can upgrade beta releases - they will only update previous normal releases. Consequently to get from a beta release to the normal release when it comes out requires a re-install and you typically lose all your data and config. A time bomb is a very reasonable way of ensuring those betas don't live forever in the hands of people who are reluctant to start over with the normal release.

    One piece of software I worked on only did the timebomb check at startup for the evaluation versions. It ran as a UNIX based service and the timebomb was for 30 days. Needless to say we ended up with people running for 6 to 12 months before rebooting their machines/power failure etc and then finding our software no longer worked. It made for very interesting sales conversations.
  • How about time bombs in production software? A partner wanted me to put a time bomb in because the customer had 30 days to pay the balance after installation of his "system". So sure enough, he forgot to pay until the last minute and we had to rush a release to him to keep it going. The whole thing was silly and we it made the customer think we didn't trust him.
Page 1 of 2 (24 items) 12