Lots of users? Some corporate, some personal, some free? POP, IMAP, and webmail? High uptime? Sounds like Scalix.
OpenMail (on which Scalix was based) scaled to insane levels compared with Exchange, Scalix should be the same. If we're talking consumer ISP-style workloads, you should be able to approach 100K users on a smallish Intel server. The key is to have a decent SAN, as previous posters have pointed out.
Scalix can support just about every Outlook feature that Exchange can (forms being the notable exception). Any mailbox can be used with POP, IMAP, Outlook/MAPI, or the Scalix web client (SWA). SWA is an AJAX client, with a look'n'feel close to Outlook.
Scalix quotes 99.99% uptime, and I saw even better in OpenMail days. Again, a good SAN is a must.
Yes and no. Most of these use a bizarre "connector" that sits on the sidelines and syncs with an Outlook PST (personal store). Very fragile architecture.
Scalix for one uses a "true" MAPI service provider, which essentially makes Outlook think it's talking to an Exchange box.
I know too much about this stuff, 'cos I used to work on HP OpenMail for like 13 years...
Scalix is probably the most fully-featured. As far as I know, the only feature it's "missing" is Outlook forms, which has been deprecated by Microsoft for a very long time.
Sigh. This is an alphaWorks project that's been kicking around for a while. Precis: it tries to match the sender IP to the purported sender domain. If it can't find a match, it falls back to something similar to challenge/response. The theory goes:
1. All spam is spoofed, so it will fail the IP/domain match and won't get past the challenge.
2. The vast majority of legitimate mail will pass the IP/domain match, so will be delivered without needing a challenge.
3. The only legitimate mail that needs to be challenged is sent by "power" users, who will know how to deal with a challenge.
This could initially cause false positive problems for some legitimate direct marketers who use some bulk email service providers. However, the problem is quite easily fixed.
Note that this doesn't fight spam, so much as fight spoofed senders. Much like SPF, in fact. Note also that there's been a deal of lousy reporting (say hello to WSJ and CNN), saying that FairUCE somehow spams the spammers back. What a load of old cobblers, as we say over here.
From the quotes attributed to an IBM exec in the WSJ, I'm worried that this mis-reporting might actually be IBM's fault.
Yes, you reminded me of my other methodology comment. This thing about putting AV in front of the spam filter.
I think this is likely to significantly reduce the effectiveness (ie, increase FNs) of *some* of the products. That very fact makes your effectiveness comparisons dubious, because you're no longer comparing apples with apples.
"Shaving a few points off" the effectiveness of all the products would be fine, but I don't see that you're shaving off equal amounts across the board. How can you be? The products are different, using different methodologies.
In order to achieve 95%+ with vanishingly-small FP rates, these guys have to make some incredibly subtle judgements. By insulating them from the out-of-band data, you're going to reduce their ability to make those judgements.
That doesn't necessarily make those products a "piece of cr@p". On the plus side, it does make them less susceptible to content-mangling tricks, which is a "zero-hour" benefit, at the very least.
I hear your point about FN/FP rates not being the only criteria, but it's absolutely Job #1 to get this right. Especially with FPs, otherwise the cure is worse than the disease.
Yes, the parent post is misguided (and yes, Joel is doing a great job with his replies here on/.), but there's a big problem with this testing methodology.
Basically (please correct me if I'm wrong, Joel), they replay a bunch of incoming messages at the product under test "in real time." This has the advantage of making the tests repeatable. However, it takes no account for the newer spam filtering methods that look at "out of band" information to see if the sender is a spamtool or a zombie.
The methodology was OK a year ago, but I have serious reservations about it now. I'm guessing this is why several of the big names declined to participate this year.
Of course, what's Joel to do? The alternative methodology is to give each product a real live inbound stream to work on, but that's hardly a repeatable test, is it?
They're already managing to do this. They have peer ing agreements with most of the DSL and cable providers here, and only allow access to some of their content to customers of those folks (who are "guaranteed" to be in the UK;-).
I can't believe nobody's mentioned Martin Lewis's Money Saving Expert yet. Highly recommended. Some of it might be interesting for folks in other countries, too.
I got into a pres conference held by Bruce Perens yesterday (actually held by his book publisher, but whatever). He had some pretty hard-hitting words to say about SCO and their motives. More so than usual, and no legalese fudging any more.
Basically, he said that SCO are committing fraud, with the objective of hiking their stock price.
There are 3 bands commonly used by GSM. Only 2 of these are used in the US, since the third is in the unlicensed 900MHz spectrum
There are actually four GSM/GPRS bands in use around the world: 850, 900, 1800, and 1900 MHz. The US uses 1900, and is beginning to use 850 (I believe freed up from old TDMA and AMPS cells). 850 is also used in South America.
The parent post implies that 1800 is also used in the US: I'm not aware that's true, but perhaps that's a new thing too.
Europe and most other countries use 900 and 1800. So, if you need to roam worldwide (with notable exceptions of CDMA-only Japan and Korea), for best flexibility you need one of the new breed of quad-band phones from Motorola or Handspring/PalmOne.
That article proves nothing of the sort. It shows a strong correlation between GI and IR. The biggest anomolies are where the GI is high and the IR is *higher*
Yes, really, old chap. For many years now, the USA have run what they call a "Visa waiver program" (sic). My US visa expired years ago, but I still enter the US for work roughly 10 times per year, simply by filling in an I-94(W) "landing card".
Last time I checked, citizens of an EU country don't need a visa for a stay in the USA of up to 90 days, even if they're working.
Of course, why should we expect that this publication should tell the Feds' side of the story, or do anything else resembling "proper" journalism, eh?;-)
Sure the US was arguably the first nation to produce a programmable electronic computer, the UK built electronic computing devices before the US, however they weren't programmable.
You miss the point: the UK built the first electronic computing devices (before the US), AND the UK also built the first programmable computing devices. ENIAC was a toy.
The different direction of slashes is due to MS-DOS using '/' as a way to denote command-line options (e.g. DEL/S/F *.* ). Early version of DOS didn't have a heirachic filesystem, so when MSFT added the concept of directories, they couldn't easily choose '/' as the separator, so they thought that they should use '\' instead.
Muscle memory sanity for people switching between DOS and Unix wasn't exactly seen as an issue to those guys;-)
It's interesting, but it doesn't appear to support much of the Outlook functionality (we went and checked with the SuSE guys at LWE last month). There's a bunch of things it just doesn't do (or you need to go to an external app. to do it). The typical Outlook-bigot HATES this approach. Samsung Contact does everything through Outlook.
Expensive? You're kidding, right? In every case I've looked at, it's way cheaper to buy than Exchange. Also, the real-life ownership costs (TCO; as measured by real users) are far, far less. Check out http://www.ferris.com/rep/200208/ for independent analysis.
Re:What can I do to show Apple that I am not happy
on
Broken .Mac?
·
· Score: 1
Slashdot Mirror
OpenMail (on which Scalix was based) scaled to insane levels compared with Exchange, Scalix should be the same. If we're talking consumer ISP-style workloads, you should be able to approach 100K users on a smallish Intel server. The key is to have a decent SAN, as previous posters have pointed out.
Scalix can support just about every Outlook feature that Exchange can (forms being the notable exception). Any mailbox can be used with POP, IMAP, Outlook/MAPI, or the Scalix web client (SWA). SWA is an AJAX client, with a look'n'feel close to Outlook.
Scalix quotes 99.99% uptime, and I saw even better in OpenMail days. Again, a good SAN is a must.
The reviewer got that very wrong. The Exchange server itself costs $4000 per box (or $700 for the crippled version). He kinda forgot that timy point.
Yes and no. Most of these use a bizarre "connector" that sits on the sidelines and syncs with an Outlook PST (personal store). Very fragile architecture.
Scalix for one uses a "true" MAPI service provider, which essentially makes Outlook think it's talking to an Exchange box.
I know too much about this stuff, 'cos I used to work on HP OpenMail for like 13 years...
Scalix is probably the most fully-featured. As far as I know, the only feature it's "missing" is Outlook forms, which has been deprecated by Microsoft for a very long time.
Untrue. The SPF presence check is just one of many tests. Much more discussion of this at m'blog.
I'm replying to my own post, 'cos I picked up a few anonymous flame comments and hate-mail. At the risk of feeding trolls, I offer this response.
Summary: It sucks. I wanted to like it, but ... I didn't.
Sigh. This is an alphaWorks project that's been kicking around for a while. Precis: it tries to match the sender IP to the purported sender domain. If it can't find a match, it falls back to something similar to challenge/response. The theory goes:
1. All spam is spoofed, so it will fail the IP/domain match and won't get past the challenge.
2. The vast majority of legitimate mail will pass the IP/domain match, so will be delivered without needing a challenge.
3. The only legitimate mail that needs to be challenged is sent by "power" users, who will know how to deal with a challenge.
This could initially cause false positive problems for some legitimate direct marketers who use some bulk email service providers. However, the problem is quite easily fixed.
Note that this doesn't fight spam, so much as fight spoofed senders. Much like SPF, in fact.
Note also that there's been a deal of lousy reporting (say hello to WSJ and CNN), saying that FairUCE somehow spams the spammers back. What a load of old cobblers, as we say over here.
From the quotes attributed to an IBM exec in the WSJ, I'm worried that this mis-reporting might actually be IBM's fault.
Yes, you reminded me of my other methodology comment. This thing about putting AV in front of the spam filter.
I think this is likely to significantly reduce the effectiveness (ie, increase FNs) of *some* of the products. That very fact makes your effectiveness comparisons dubious, because you're no longer comparing apples with apples.
"Shaving a few points off" the effectiveness of all the products would be fine, but I don't see that you're shaving off equal amounts across the board. How can you be? The products are different, using different methodologies.
In order to achieve 95%+ with vanishingly-small FP rates, these guys have to make some incredibly subtle judgements. By insulating them from the out-of-band data, you're going to reduce their ability to make those judgements.
That doesn't necessarily make those products a "piece of cr@p". On the plus side, it does make them less susceptible to content-mangling tricks, which is a "zero-hour" benefit, at the very least.
I hear your point about FN/FP rates not being the only criteria, but it's absolutely Job #1 to get this right. Especially with FPs, otherwise the cure is worse than the disease.
richi.
Yes, the parent post is misguided (and yes, Joel is doing a great job with his replies here on /.), but there's a big problem with this testing methodology.
Basically (please correct me if I'm wrong, Joel), they replay a bunch of incoming messages at the product under test "in real time." This has the advantage of making the tests repeatable. However, it takes no account for the newer spam filtering methods that look at "out of band" information to see if the sender is a spamtool or a zombie.
The methodology was OK a year ago, but I have serious reservations about it now. I'm guessing this is why several of the big names declined to participate this year.
Of course, what's Joel to do? The alternative methodology is to give each product a real live inbound stream to work on, but that's hardly a repeatable test, is it?
r.
They're already managing to do this. They have peer ing agreements with most of the DSL and cable providers here, and only allow access to some of their content to customers of those folks (who are "guaranteed" to be in the UK ;-).
r.
r.
Utter tripe. For one, we don't say "tube" in the UK (UoSAT-2 is British).
I got into a pres conference held by Bruce Perens yesterday (actually held by his book publisher, but whatever). He had some pretty hard-hitting words to say about SCO and their motives. More so than usual, and no legalese fudging any more.
Basically, he said that SCO are committing fraud, with the objective of hiking their stock price.
You're looking for DSNs (delivery service notifications; RFCs 3461 and 3464) and MDNs (message disposition notifications; RFC 3503).
richi.
There are actually four GSM/GPRS bands in use around the world: 850, 900, 1800, and 1900 MHz. The US uses 1900, and is beginning to use 850 (I believe freed up from old TDMA and AMPS cells). 850 is also used in South America.
The parent post implies that 1800 is also used in the US: I'm not aware that's true, but perhaps that's a new thing too.
Europe and most other countries use 900 and 1800. So, if you need to roam worldwide (with notable exceptions of CDMA-only Japan and Korea), for best flexibility you need one of the new breed of quad-band phones from Motorola or Handspring/PalmOne.
richi.
That article proves nothing of the sort. It shows a strong correlation between GI and IR. The biggest anomolies are where the GI is high and the IR is *higher*
r.
Yes, really, old chap. For many years now, the USA have run what they call a "Visa waiver program" (sic). My US visa expired years ago, but I still enter the US for work roughly 10 times per year, simply by filling in an I-94(W) "landing card".
r.
Last time I checked, citizens of an EU country don't need a visa for a stay in the USA of up to 90 days, even if they're working.
;-)
Of course, why should we expect that this publication should tell the Feds' side of the story, or do anything else resembling "proper" journalism, eh?
r.
You miss the point: the UK built the first electronic computing devices (before the US), AND the UK also built the first programmable computing devices. ENIAC was a toy.
r.
The different direction of slashes is due to MS-DOS using '/' as a way to denote command-line options (e.g. DEL /S/F *.* ). Early version of DOS didn't have a heirachic filesystem, so when MSFT added the concept of directories, they couldn't easily choose '/' as the separator, so they thought that they should use '\' instead.
;-)
Muscle memory sanity for people switching between DOS and Unix wasn't exactly seen as an issue to those guys
r.
It's interesting, but it doesn't appear to support much of the Outlook functionality (we went and checked with the SuSE guys at LWE last month). There's a bunch of things it just doesn't do (or you need to go to an external app. to do it). The typical Outlook-bigot HATES this approach. Samsung Contact does everything through Outlook.
Expensive? You're kidding, right? In every case I've looked at, it's way cheaper to buy than Exchange. Also, the real-life ownership costs (TCO; as measured by real users) are far, far less. Check out http://www.ferris.com/rep/200208/ for independent analysis.
> What can I do to show Apple that I am not happy
/. ?
Ummm, post to
Your bootable partition shouldn't be bigger than 8GB.