Lotus lashes out over MS Exchange v Domino test
Now, where have we heard this one before?
What had been a little local skirmish between Microsoft and Lotus over Notes may be escalating into a war, despite the fixed smiles of chumminess following Lotus' decision this week to offer MS Outlook with Notes. Microsoft had of course hoped that Notes would fall to Exchange as had 1-2-3 fallen to Excel many years earlier, but it didn't happen. From time to time, Microsoft ran a campaign to show that "independent research" purported to prove that Exchange Server was "preferred by IT administrators", was "easier, faster than Lotus Domino" or that it "outperformed" Domino. The skirmishes were punctuated by announcements that Lotus and Microsoft would "promote interoperability on the Web" with IE4 (July 1997), or integrate Windows Media Technologies into Lotus Notes and Domino" (last September). The recent trouble started on 10 January when Microsoft announced that "an independent laboratory test" showed that "Exchange leads in email performance". It turned out that the lab was ZD Labs. After disclaiming any responsibility for "the accuracy and validity of its testing... test results and analysis, their accuracy completeness or quality" (although "reasonable efforts" had been made), the ZD report went on to make what Lotus claimed were serious blunders; in a response just published, the company says that Microsoft's "press release is misleading, and the tested Domino configuration is flawed". The details of five "inaccurate statements" in Microsoft's press release are documented in the Lotus response, but a further accusation is that an unidentified "leading multinational consulting firm" (can any Register reader tell us who?) hired to tune the Domino server did not understand what it was doing. No prizes for guessing who suggested that Exchange be set up with "the operating system, swap and binaries on one drive set and the data on the other". But the Domino test was done without consulting Lotus, which says the sub-optimal setting up the "operating system, binaries (Domino executables), swap file, and Domino data on the same controller, and the last two on the same drive set" was pretty certain to ensure that Exchange won. Microsoft tried to get Lotus to endorse the study afterwards, but Lotus wanted a broader test to which Microsoft would not agree. Reading between the lines, it looks as though Microsoft had found one particular setup in which Exchange either beat Domino or came close to it, and that it was not about to surrender this straw. There were some barbs in the Lotus response: although Microsoft had claimed that "all tests were done using SMTP, POP and IMAP protocols" the testing excluded IMAP, evidently because Exchange has "a limit of 2,500 users per server". Lotus also pointed out that Microsoft has stopped claiming that Exchange is an "email and collaboration server", and calls it just an email server that has as its main feature "a transacted database for full recovery". That's not all however. Lotus pointed out that Exchange is usually configured with 250 to 500 mailboxes per server (or less - the Merrill Lynch Exchange 5.5 deployment has 70,000 users with 700 servers). Lotus also noted that users should be sceptical of any claim for 5,000 or more users on a single server, and that anybody seeking a reference site should email Mike Tuchen, the Microsoft product manager who sponsored the study. ®