Y2.01K bug trips up Symantec

Symantec's Endpoint Protection Manager has been hit by a classic date bug and fell over at the end of the year, accepting no definition updates dated since then. Symantec have issued a statement, which states that: "An issue has been identified in the Symantec Endpoint Protection Manager (SEPM) server whereby all types of SEP …

[Ken Hagan]

if <10 = 20xx

1911-2010 is a very curious window, particularly for time-stamping the output of a software system that presumably didn't exist for the first 75 years of that same window.

The cause is more likely to be a single digit date field. Perhaps this was a decision made in the early nineties and papered over before January 2000 (I can't find any evidence that Symantec were similarly embarrassed ten years ago) but not properly fixed.

[Anonymous Coward]

Google "Symantec Sucks"

Google the search term: "Symantec Sucks". It's clear that they're idiots.

[Anonymous Coward]

@Ken

You need a low date threshold in some software as it deals with dates of birth (but not so with anti virus software).

10 years is a long time. I've changed company twice since then. Always a good idea to do if you have made such a timebomb/lazy change.

[Dale Richards]

Uh-oh

We're running SEP here - current definitions are at 31 December 2009 r116. What happens when they get past r999?

PANIC!

[Anonymous Coward]

@Dale Richards

31 December 2009 r1000, perhaps?

[Ken Hagan]

Re: @Dale Richards

Perhap, but the only thing we know for certain is that the original programmers failed to allow enough space in their data formats. Given that the article mentions revision "041", it would be a brave soul who assumed that the revision isn't a three digit field, making r1000 unlikely. It could even be just the zero justified decimal representation of a single byte, in which case r256 looks unlikely, too.

[Rhyd]

Perhaps...

p000?

[Anonymous Coward]

all those smug asses

who said the Y2K bug never happened can now shut up and fix something

[Michael C]

Dumb.

They were not testing actively dates into the next callendar year at all? ANY proper dev team does this every year, if nothing more than to validate leap year, daylight savings changes, and more.

It's no wonder we shifted off Symantec productcs entirely many years ago, and have yet not gone back.

[TB]

10 years on and same Symantec problem

Anyone remember "WinFax pro"?

In 1999 I checked with Symantec that it was Win2K safe and was assured it was OK. The new millennium came and all was well for 2 months, then at the end of February it collapsed, with date problems. Did they ever offer a fix? Well if so I never received one and so I moved on and vowed never to use Symantec software again, which now seems to have been a shrewd move.

--

TB

[Anonymous Coward]

What will be next? Windows?

Yarrt - so glad my company is off Symantec Antivirus products.

Didn't Symantec's QA department tested their software fully before releasing it?

Wonder which software will go bonkers next? Seems that Facebook also do have date issues...

[Alice Andretti]

"Symantec" and "quality" in the same sentence...

Their quality has been MIA for well over a decade. :( That also includes a recent interval where I foolishly gave them a second chance, this time with their Windows version of stuff only to end up ditching Norton again, this time permanently.

Yes, yes, I know, no software is perfect, but sometimes it seems that Symantec isn't even trying. Maybe they figure the typical grandpa Norton user (or whatever) can't tell the difference, who knows.

Does Symantec not pay their employees enough to get quality work, or what the hell *is* the problem there anyway?

[Anonymous Coward]

Norton

Used Norton for years (company policy) and coud never understand why people hated them so much. Now we've upgraded to Symantec Endpoint Protection, and light has dawned. Symantec Sucks, big time!

[typeo]

Products on a downward slope

SEP has to be their worst product yet. Released with bugs from the outset and each time a new build is released you have to perform an upgrade of the servers or clients not just a simple Live Update. And the overhead of the client software seems to get worse with each release.

From past experiences no AV vendor is without faults but Symantec is king of the mountain for the number of faults with its software.

Symantec have also done a damn fine job of messing up Veritas products since they have taken them over. The support is awful.

[TroubleAhead]

SEP 11 failed Eicar - Maybe you should scan Event logs

We have SEP 11 MR5. Today one of our servers failed an Eicar test - I could unzip and run the eicar.com file without any AV warning or intervention - it should stop the file being unzipped. This is the most basic test for any AV product.

The SEP client does not report any issues, the central console does not report any issues. We have had an issue before and have been scanning event logs for event ID 4:

Event Type: Error

Event Source: SRTSP

Description: Error loading virus definitions.

Not all servers with this error fail an Eicar. If you are running SEP 11, maybe you should dust off EventComb and run a search.... now.

[Henry Wertz 1]

Norton must be...

Norton must be turning over in his grave.. well, if he's dead. Otherwise I wonder if he's pretty pissed. The *old* norton utilities (Peter Norton and all that) were great, they were small, fast, and each utility did what it claimed without bugging out. They didn't IMMEDIATELY go downhill once Symantec bought them up, but they got seemingly a little bigger, buggier, and slower with every release, it's been ridiculous for at least 10 years.