Обсуждение: Re: [GENERAL] Oracle buys Innobase
-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 > I don't think that PostgreSQL is really on Oracle's radar at the moment. Please don't make this assumption. PostgreSQL is *very* much on their radar, and probably represents the biggest long-term threat to their core database business at the moment. We got a hint of that during the .org bidding, but for now it is in Oracle's interest not to call attention to PostgreSQL. The last thing they want is publicity for the project. We may be a harder target to hurt than MySQL, but we are a target, make no mistake about it. I'm sure PostgreSQL is on the radar of Sybase, Microsoft, and IBM as well. - -- Greg Sabino Mullane greg@turnstep.com PGP Key: 0x14964AC8 200510170838 http://biglumber.com/x/web?pk=2529DF6AB8F79407E94445B4BC9B906714964AC8 -----BEGIN PGP SIGNATURE----- iD8DBQFDU5wKvJuQZxSWSsgRAryKAKCkK1+4raUv0MOr/TaFOJoV6Vm9sACfXHDC fRH5+YBEVRQzih/yQYr6Su4= =AX08 -----END PGP SIGNATURE-----
> Please don't make this assumption. PostgreSQL is *very* much on their radar, > and probably represents the biggest long-term threat to their core database > business at the moment. We got a hint of that during the .org bidding, but > for now it is in Oracle's interest not to call attention to PostgreSQL. > The last thing they want is publicity for the project. We may be a harder > target to hurt than MySQL, but we are a target, make no mistake about it. > I'm sure PostgreSQL is on the radar of Sybase, Microsoft, and IBM as well. And they probably read every word we write ;) Chris
On Mon, 2005-10-17 at 22:46 +0800, Christopher Kings-Lynne wrote: > And they probably read every word we write ;) ...and it will certainly slow them down :-)
Scott Marlowe wrote: >I'd bet they read plenty, but don't necessarily understand a lot, >judging by their pitiful fud campaign when Afilias proposed using >postgresql as a database behind .org. They tried to say PostgreSQL >didn't support transactions. So, while we may be on their screens, and >I'm sure some marketeer there tries to keep up with some of the traffic >here, the actual comprehension seems pretty low judging by their past >statements. > >Actually, I kinda hope it stays that way. > > Ok. I should have said "serious target." MySQL has been a serious target for a number of years. I think we are still the unknown bugaboo to them. I.e. I see no evidence that Oracle is taking the PostgreSQL threat seriously, and the FUD campaign is more evidence that they don't (there are plenty of areas where Oracle has an edge over PostgreSQL-- the idea that "PostgreSQL doesn't support transactions" can only indicate that this was a cursory and hasty attack and maybe even a wakeup call for them, or maybe they got us mixed up with MySQL w/MyISAM). The real question is whether after the .org campaign occurred, we are now a higher-profile target that is taken more seriously. Personally, I would doubt it for reasons mentioned below. The thing is, we may be a head-to-head competitor with Oracle in many areas, but we are pretty minor compared to Sybase, Microsoft, and IBM at the moment. I.e. while we are an emerging threat, Oracle has plenty of clear and present threats to its market share to deal with. Therefore, I am willing to bet that we are probably a distant target, somewhere after Ingress II and maybe even Firebird/Interbase. This is based on the assumption that in any significantly large corporation, there will be a lot of legacy competitive effort and that the rampup time to look at new threats is really pretty large. I.e. at Microsoft when I left (2003), Java and Sun were still higher competitive priorities than Linux (and still very much in a middle-phase). From what I have read after leaving, I think that Microsoft's strategy is still in an opening phase mostly consisting of GetTheFUD and internal product research. MySQL is different. They established a large user base early on, and people have a tendency (wrongly) to think of them as The Open Source RDBMS. So I am willing to be that Oracle has been ramping up a competitive strategy against them for at least five years (they showed a clear competitive strategy against them as early as 2000). The fact that they are an easier target complicates matters for them, but I think that this is more of a transition to an end-game strategy by Oracle than anything else. I will be worried if and when Oracle demonstrates any intelligent competitive strategy against us. A poorly orchestrated and hasty FUD campaign does not qualify. Best Wishes, Chris Travers Metatron Technology Consulting
On Mon, 2005-10-17 at 09:46, Christopher Kings-Lynne wrote: > > Please don't make this assumption. PostgreSQL is *very* much on their radar, > > and probably represents the biggest long-term threat to their core database > > business at the moment. We got a hint of that during the .org bidding, but > > for now it is in Oracle's interest not to call attention to PostgreSQL. > > The last thing they want is publicity for the project. We may be a harder > > target to hurt than MySQL, but we are a target, make no mistake about it. > > I'm sure PostgreSQL is on the radar of Sybase, Microsoft, and IBM as well. > > And they probably read every word we write ;) I'd bet they read plenty, but don't necessarily understand a lot, judging by their pitiful fud campaign when Afilias proposed using postgresql as a database behind .org. They tried to say PostgreSQL didn't support transactions. So, while we may be on their screens, and I'm sure some marketeer there tries to keep up with some of the traffic here, the actual comprehension seems pretty low judging by their past statements. Actually, I kinda hope it stays that way.
On Monday 17 October 2005 13:01, Scott Marlowe wrote: > On Mon, 2005-10-17 at 09:46, Christopher Kings-Lynne wrote: > > > Please don't make this assumption. PostgreSQL is *very* much on their > > > radar, and probably represents the biggest long-term threat to their > > > core database business at the moment. We got a hint of that during the > > > .org bidding, but for now it is in Oracle's interest not to call > > > attention to PostgreSQL. The last thing they want is publicity for the > > > project. We may be a harder target to hurt than MySQL, but we are a > > > target, make no mistake about it. I'm sure PostgreSQL is on the radar > > > of Sybase, Microsoft, and IBM as well. > > > > And they probably read every word we write ;) > > I'd bet they read plenty, but don't necessarily understand a lot, > judging by their pitiful fud campaign when Afilias proposed using > postgresql as a database behind .org. They tried to say PostgreSQL > didn't support transactions. So, while we may be on their screens, and > I'm sure some marketeer there tries to keep up with some of the traffic > here, the actual comprehension seems pretty low judging by their past > statements. > > Actually, I kinda hope it stays that way. > Don't bet on it. If Afilias is 4 years smarter about postgresql, you can bet Oracle is too. In fact my guess is that they started reading up as soon as .org was awarded to a pg based company. I think before that they probably figured that my$ql, being more popular, was roughly equal if not better than postgresql, and often confused the two. If there smart enough to be buying innobase these days, you can bet that by now they have this stuff all straightened out. -- Robert Treat Build A Brighter Lamp :: Linux Apache {middleware} PostgreSQL
Robert Treat <xzilla@users.sourceforge.net> writes: > If there smart enough to > be buying innobase these days, you can bet that by now they have this stuff > all straightened out. No, that doesn't seem to follow ... if Oracle are spending their resources to attack MySQL rather than us, the conclusion would be that they are clearly still more informed by "the buzz" than technical merit. What seems likely to me is that the Innobase purchase was a target of opportunity --- they saw a chance to destroy a potential threat, and took it. This proves nothing about their assessment of the relative risks from us and MySQL ... only that they haven't yet thought of an equally painless way to destroy us. regards, tom lane
>>If there smart enough to >>be buying innobase these days, you can bet that by now they have this stuff >>all straightened out. > > > No, that doesn't seem to follow ... if Oracle are spending their > resources to attack MySQL rather than us, the conclusion would be that > they are clearly still more informed by "the buzz" than technical merit. With no disrespect to PostgreSQL, MySQL has 100x our downloads and installations... Oracle is simply going after by far the biggest open source database player... Chris
On Tuesday 18 October 2005 23:44, Chris Travers wrote: > Christopher Kings-Lynne wrote: > > With no disrespect to PostgreSQL, MySQL has 100x our downloads and > > installations... > > > > Oracle is simply going after by far the biggest open source database > > player... > > As I said, Oracle demonstrated in 2000 that they had already singled > MySQL out for special competitive treatement. They did this by starting > to offer db conversion utilities in order to help people migrate from > MySQL to Oracle. It is not about technical merit, it is about market > share. We could have the best RDBMS in the world but if we never get wadda ya mean "could"?" :-) > enough users to directly threaten them to the level that MS SQL Server > or DB2 does, we are not the threat that they are, and we are not worth > the time and expense that research, competitive strategizing, etc. would > incur. Therefore, I suspect that we are sort of on the back burner > competitive strategy wise. I.e. competition is on a project-by-project > basis, and not coordinated as of yet. > > There are some things on the horizon that could change this quite > quickly, however: > > 1) Sun is talking about packaging PostgreSQL and distributing it with > Solaris. This would bring us directly head to head with Oracle in a > large number of potential installations. > > 2) EnterpriseDB's efforts and awards may have attracted some > attention. This may reinforce the idea that we are a threat. > > If this is the case, I bet that Oracle is probably pressuring Sun not to > distribute PostgreSQL, and if they do anyway, we need to be concerned > about the beginning of a high-level coordinated strategy targetting us > specifically. IMO, it is likely to start with one of two things: > > 1) PostgreSQL to Oracle database conversion utilities released by > Oracle (unlikely given extensible languages in PostgreSQL). they need to "reverse" engineer enterprisedb :-) > 2) Some sort of FUD campaign on the part of Oracle directed > specifically at us and not tied to any specific project (fairly likely). > look for pointers to lack of benchmarks, patent issues, and great bridge... those seem to be the most common rehash of fud. -- Robert Treat Build A Brighter Lamp :: Linux Apache {middleware} PostgreSQL
Chris Travers <chris@verkiel.metatrontech.com> writes: > IMO, it is likely to start with one of two things: > 1) PostgreSQL to Oracle database conversion utilities released by > Oracle (unlikely given extensible languages in PostgreSQL). > 2) Some sort of FUD campaign on the part of Oracle directed > specifically at us and not tied to any specific project (fairly likely). Well, #1 would require quite a nontrivial investment of time by Oracle (I doubt they'd even think about offering ports of any PL other than plpgsql, and still it'd be a major project). #2 only requires inventing some plausible lies. So you can bet we'll see #2 long before #1. As Andrew noted, we've already heard plenty of FUD from Oracle. What we've not seen is a FUD campaign based on serious study of our weaknesses --- they've only bothered to muster transparent attacks on "open source DBs" in general. My prediction is that the next step will be FUD that's really designed specifically against Postgres. regards, tom lane
> 1) PostgreSQL to Oracle database conversion utilities released by > Oracle (unlikely given extensible languages in PostgreSQL). Strangely a pgsql to oracle exporter is a good thing. It'd be a great feature of PostgreSQL. Imagine how many people would start on PostgreSQL if they KNEW that one day they could easily move to Oracle if they needed to. Risk management. Chris
> As Andrew noted, we've already heard plenty of FUD from Oracle. What > we've not seen is a FUD campaign based on serious study of our > weaknesses --- they've only bothered to muster transparent attacks on > "open source DBs" in general. My prediction is that the next step will > be FUD that's really designed specifically against Postgres. I admit I must have missed all this '.org FUD' - is it still around. I really don't know what you guys are referring to. Chris
Christopher Kings-Lynne <chriskl@familyhealth.com.au> writes: > Strangely a pgsql to oracle exporter is a good thing. It'd be a great > feature of PostgreSQL. Imagine how many people would start on > PostgreSQL if they KNEW that one day they could easily move to Oracle if > they needed to. Risk management. Problem is: to offer such a thing with a straight face, we'd have to confine ourselves to an Oracle-subset version of SQL. For instance, lose the ability to distinguish empty-string from NULL. regards, tom lane
Tom Lane wrote: >Christopher Kings-Lynne <chriskl@familyhealth.com.au> writes: > > >>Strangely a pgsql to oracle exporter is a good thing. It'd be a great >>feature of PostgreSQL. Imagine how many people would start on >>PostgreSQL if they KNEW that one day they could easily move to Oracle if >>they needed to. Risk management. >> >> > >Problem is: to offer such a thing with a straight face, we'd have to >confine ourselves to an Oracle-subset version of SQL. For instance, >lose the ability to distinguish empty-string from NULL. > > Funny how many Oracle DBA's don't seem to be aware of this "feature" .... Funny how many of those that are seem to think that this is perfectly OK :-( On the other hand, it is one weakness of Oracle that has not gotten a major airing in public... It might be worth saving for when the FUD starts to fly :-) Best Wishes, Chris Travers Metatron technology Consulting
> Problem is: to offer such a thing with a straight face, we'd have to > confine ourselves to an Oracle-subset version of SQL. For instance, > lose the ability to distinguish empty-string from NULL. I wasn't saying we write it - let Oracle do it :D Chris
On Wed, 19 Oct 2005, Christopher Kings-Lynne wrote: >>> If there smart enough to be buying innobase these days, you can bet that >>> by now they have this stuff all straightened out. >> >> >> No, that doesn't seem to follow ... if Oracle are spending their >> resources to attack MySQL rather than us, the conclusion would be that >> they are clearly still more informed by "the buzz" than technical merit. > > With no disrespect to PostgreSQL, MySQL has 100x our downloads and > installations... Yeah, kids playing with toys. You can't imagine how many people I heard have MySQL installed in there win98. The bad thing is they addopt MySQL because they could have it installed there. :-( -- 11:55:01 up 155 days, 1:49, 3 users, load average: 0.12, 0.72, 0.86 ----------------------------------------------------------------- Lic. Martín Marqués | select 'mmarques' || '@' || 'unl.edu.ar' Centro de Telematica | DBA, Programador, Administrador Universidad Nacional del Litoral -----------------------------------------------------------------
On Wed, 2005-10-19 at 12:51 +0200, Martijn van Oosterhout wrote: > Just for the hell of it I looked at the popcon stats for debian > installs (see below). It tells me the following: > > - Something like half the people who install mysql-server (any version) > never use it. People who install PostgreSQL are (slightly) more likely > to actually use it. > > - For mysql, users of the client are approximatly twice the amount that > use the server. For postgres, the client and server count is about the > same. This one is curious, don't know what to make of it. When you install Debian from scratch, the tasksel list offers you the chance to install database packages. If you select that, it installs postgresql server rather than mysql, which may help the statistics in Debian. The postgresql server package depends on postgresql-client. I think that the only people to install postgresql-client without the server would be those with multiple machines communicating with a server and a number of those might install the server by mistake. The ratio of nearly 6 to 4 seems quite reasonable. -- Oliver Elphick olly@lfix.co.uk Isle of Wight http://www.lfix.co.uk/oliver GPG: 1024D/A54310EA 92C8 39E7 280E 3631 3F0E 1EC0 5664 7A2F A543 10EA ======================================== Do you want to know God? http://www.lfix.co.uk/knowing_god.html
I think this probably belongs back on -advocacy, so I'm cc:ing there so we can move it. On Tue, Oct 18, 2005 at 03:16:23PM -0700, Chris Travers wrote: > Interesting. So they are willing to appear ill-informed in public but > better informed in private? To what end? That seems strange to me.... To the end of dismissing the serious-but-free competition in public. If Oracle is talking to the computer press, they have enough experience to know just how much they can play with stating the way the world is, and have it quoted verbatim as revealed truth. Apart from database weenies like us, people reading the Oracle pronouncement conflating PostgreSQL and other database systems will just think it's true. After all, Oracle said it, and the press guy from InfoWorld must have checked it out, right? If you think I'm being unduly cynical, note that the Gartner comments in their consulting for ICANN in the .org reassignment basically argued that PostgreSQL was a significant risk because it wasn't Oracle. There's nothing _wrong_ with that way of thinking -- corporations are mostly about stability, which means following conventional (==safe) wisdom. But that mindset is something that Oracle is skilled at exploiting, and I'm not surprised they do it against PostgreSQL (even if their behaviour sounds irrational to someone who really knows the capabilities of the various systems). But that isn't really why I replied to this :) > This is one reason why I would like to see some of us push PostgreSQL > into a role of *the* RDBMS to study for RDBMS theory. Unfortunately > this means a lot of documentation written by experts interested in > really teaching beginners the right way to do things.... I don't > consider myself qualified to do this by myself. I like this idea. I wonder how to get it moving. A -- Andrew Sullivan | ajs@crankycanuck.ca "The year's penultimate month" is not in truth a good way of saying November. --H.W. Fowler
On Wednesday 19 October 2005 09:58, Andrew Sullivan wrote: > > This is one reason why I would like to see some of us push PostgreSQL > > into a role of *the* RDBMS to study for RDBMS theory. Unfortunately > > this means a lot of documentation written by experts interested in > > really teaching beginners the right way to do things.... I don't > > consider myself qualified to do this by myself. > > I like this idea. I wonder how to get it moving. Support those companies that do training, hint hint ;-)
On Wed, Oct 19, 2005 at 01:11:10PM +0800, Christopher Kings-Lynne wrote: > I admit I must have missed all this '.org FUD' - is it still around. I > really don't know what you guys are referring to. I mentioned this in another message, but all the redelegation stuff is available from ICANN's site. The Oracle comments were in this thread: http://forum.icann.org/org-eval/gartner-report/ A -- Andrew Sullivan | ajs@crankycanuck.ca This work was visionary and imaginative, and goes to show that visionary and imaginative work need not end up well. --Dennis Ritchie
Andrew Sullivan wrote: >I think this probably belongs back on -advocacy, so I'm cc:ing there >so we can move it. > >On Tue, Oct 18, 2005 at 03:16:23PM -0700, Chris Travers wrote: > > >>Interesting. So they are willing to appear ill-informed in public but >>better informed in private? To what end? That seems strange to me.... >> >> > >To the end of dismissing the serious-but-free competition in public. >If Oracle is talking to the computer press, they have enough >experience to know just how much they can play with stating the way >the world is, and have it quoted verbatim as revealed truth. Apart >from database weenies like us, people reading the Oracle >pronouncement conflating PostgreSQL and other database systems will >just think it's true. > You mean like that RDBMS comparison on Dev-X by the marketing head of Daffodil? Somehow I don't think that this is limited to Oracle... > After all, Oracle said it, and the press guy >from InfoWorld must have checked it out, right? If you think I'm >being unduly cynical, note that the Gartner comments in their >consulting for ICANN in the .org reassignment basically argued that >PostgreSQL was a significant risk because it wasn't Oracle. There's > > >nothing _wrong_ with that way of thinking -- corporations are mostly >about stability, which means following conventional (==safe) wisdom. > > Sure. But in this case, Gartner came across as poorly informed. I think that this sort of strategy can only go on so long before people assume that Gartner is largely acting as paid shills, and Gartner looses a bunch of buisness. You have to understand how this process likely works. ICANN hires Gartner. Gartner asks Oracle what they think. Oracle directs the question to their PR agents, who then ask if anyone has any information. The PR agents who don't understand the technology are then in charge of repackaging the data up to suit Oracle's agenda. They probably then do some analysis on this data, maybe ask a few of Oracle's customers what they think, and call it a day. I think that this happens because the industry is far to vast for anyone to really know all of it well, especially as an industry analyst. So these firms get their data from the big, entrenched companies. However, I think that these firms will suffer huge reputation issues as the industry changes. >But that mindset is something that Oracle is skilled at exploiting, >and I'm not surprised they do it against PostgreSQL (even if their >behaviour sounds irrational to someone who really knows the >capabilities of the various systems). > > I am sure that they do. But they will do the same against any other RDBMS. In other words, if Oracle has formed an internal team to focus on "how to beat PostgreSQL" I don't think they have done it very long ago. >But that isn't really why I replied to this :) > > > >>This is one reason why I would like to see some of us push PostgreSQL >>into a role of *the* RDBMS to study for RDBMS theory. Unfortunately >>this means a lot of documentation written by experts interested in >>really teaching beginners the right way to do things.... I don't >>consider myself qualified to do this by myself. >> >> > >I like this idea. I wonder how to get it moving. > >A > >
>> >> To the end of dismissing the serious-but-free competition in public. >> If Oracle is talking to the computer press, they have enough >> experience to know just how much they can play with stating the way >> the world is, and have it quoted verbatim as revealed truth. Apart >> from database weenies like us, people reading the Oracle >> pronouncement conflating PostgreSQL and other database systems will >> just think it's true. >> > You mean like that RDBMS comparison on Dev-X by the marketing head of > Daffodil? Somehow I don't think that this is limited to Oracle... FYI, I have been tasked by Devx to rewrite that comparison. Sincerely, Joshua D. Drake -- The PostgreSQL Company - Command Prompt, Inc. 1.503.667.4564 PostgreSQL Replication, Consulting, Custom Development, 24x7 support Managed Services, Shared and Dedicated Hosting Co-Authors: plPHP, plPerlNG - http://www.commandprompt.com/
On Wed, Oct 19, 2005 at 02:25:58PM -0700, Chris Travers wrote: > Oracle's customers what they think, and call it a day. I think that > this happens because the industry is far to vast for anyone to really > know all of it well, especially as an industry analyst. So these firms > get their data from the big, entrenched companies. However, I think > that these firms will suffer huge reputation issues as the industry changes. We're going to have to agree to disagree about this; but I don't think for a second that markets are actually efficient at driving out nonsense and cant. If they were, the current stock markets wouldn't work, and InfoWorld would have gone out of publication ages ago. The computer industry has been like this practically forever: in the past, there was a safe choice (IBM), a moderately safe choice (the 7 Dwarves), and something that would get you fired if it didn't work. And consulting companies exist exactly to parrot whatever the current conventional wisdom is. Gartner only came across as ill-informed in that report (or any other I've ever read by them) when I actually knew something about the technolgies in question. Do I trust anything that Gartner says? No. But I'm not their market, and never will be. Their market is conventional-grade MBAs who have never studied anything except commerce. Those people want affirmation of their prejudices; not truth. Their prejudices will change over time, but only if we are successful in moving PostgreSQL from the "get you fired if it doesn't work" into the "Seven Dwarves" category. A -- Andrew Sullivan | ajs@crankycanuck.ca I remember when computers were frustrating because they *did* exactly what you told them to. That actually seems sort of quaint now. --J.D. Baldwin
Andrew Sullivan wrote: > And consulting companies exist exactly to parrot whatever the current > conventional wisdom is. Gartner only came across as ill-informed in > that report (or any other I've ever read by them) when I actually > knew something about the technolgies in question. Gartner seem to have the same problems as the wider press do with any technical field - they don't know what they're talking about. When it is reporting on who said what when, both are fine. When it's discussing figures, they seem broadly OK. Once you need to get technical they generally flounder. -- Richard Huxton Archonet Ltd
Christopher Kings-Lynne wrote: > With no disrespect to PostgreSQL, MySQL has 100x our downloads and > installations... > > Oracle is simply going after by far the biggest open source database > player... > As I said, Oracle demonstrated in 2000 that they had already singled MySQL out for special competitive treatement. They did this by starting to offer db conversion utilities in order to help people migrate from MySQL to Oracle. It is not about technical merit, it is about market share. We could have the best RDBMS in the world but if we never get enough users to directly threaten them to the level that MS SQL Server or DB2 does, we are not the threat that they are, and we are not worth the time and expense that research, competitive strategizing, etc. would incur. Therefore, I suspect that we are sort of on the back burner competitive strategy wise. I.e. competition is on a project-by-project basis, and not coordinated as of yet. There are some things on the horizon that could change this quite quickly, however: 1) Sun is talking about packaging PostgreSQL and distributing it with Solaris. This would bring us directly head to head with Oracle in a large number of potential installations. 2) EnterpriseDB's efforts and awards may have attracted some attention. This may reinforce the idea that we are a threat. If this is the case, I bet that Oracle is probably pressuring Sun not to distribute PostgreSQL, and if they do anyway, we need to be concerned about the beginning of a high-level coordinated strategy targetting us specifically. IMO, it is likely to start with one of two things: 1) PostgreSQL to Oracle database conversion utilities released by Oracle (unlikely given extensible languages in PostgreSQL). 2) Some sort of FUD campaign on the part of Oracle directed specifically at us and not tied to any specific project (fairly likely). Best Wishes, Chris Travers Metatron Technology Consulting
Вложения
Andrew Sullivan wrote: > > >>This is one reason why I would like to see some of us push PostgreSQL >>into a role of *the* RDBMS to study for RDBMS theory. Unfortunately >>this means a lot of documentation written by experts interested in >>really teaching beginners the right way to do things.... I don't >>consider myself qualified to do this by myself. >> >> > >I like this idea. I wonder how to get it moving. > > I think there are two aspects to this puzzle. The first is to support the companies that do training as Robert suggests. But there is a second one too. I think that we as a community would do well to create documentation about RDBMS fundamentals using PostgreSQL as the reference example. Topics might include: 1) Relational Theory 2) Normalization 3) Database Design (including Date's Central Rule) 4) Performance considerations (implimentation and non-implimentation-specific) Any others? Maybe SQL as a language? However, part of what would be required to make this work is that you would need to make the RDBMS as standards-compliant as possible, and provide the option for standard behavior even if people don't generally like one specific aspect of the standard. Best Wishes, Chris Travers Metatron Technology Consulting
Вложения
On Wed, Oct 19, 2005 at 10:55:22AM +0800, Christopher Kings-Lynne wrote: > With no disrespect to PostgreSQL, MySQL has 100x our downloads and > installations... Just for the hell of it I looked at the popcon stats for debian installs (see below). It tells me the following: - Something like half the people who install mysql-server (any version) never use it. People who install PostgreSQL are (slightly) more likely to actually use it. - For mysql, users of the client are approximatly twice the amount that use the server. For postgres, the client and server count is about the same. This one is curious, don't know what to make of it. - when it comes to client libs, a lot of people have them installed (presumably linked to various apps) but they don't apparently connect anywhere with them. Now, this is not exactly a represenative sample and statistical errors abound, and we're not counting Windows installations but 100x seems like an exaggeration to me... :) Have a nice day, #<name> is the package name; #<inst> is the number of people who installed this package; #<vote> is the number of people who use this package regularly; #<old> is the number of people who installed, but don't use this package # regularly; #<recent> is the number of people who upgraded this package recently; #<no-files> is the number of people whose entry didn't contain enough # information (atime and ctime were 0). #rank name inst vote old recent no-files (maintainer) 183 libmysqlclient12 4483 3026 663 421 373 (Christian Hammers) 266 mysql-client 2803 2188 216 172 227 (Christian Hammers) 453 libpq3 3710 1266 1065 231 1148 (Martin Pitt) 478 mysql-server 2342 1171 529 490 152 (Christian Hammers) 553 libmysqlclient14 2437 954 145 332 1006 (Christian Hammers) 583 mysql-client-4.1 1111 886 21 204 0 (Christian Hammers) 661 postgresql-client 1709 729 372 31 577 (Martin Pitt) 662 postgresql 1286 728 132 14 412 (Martin Pitt) 883 mysql-server-4.1 883 490 84 309 0 (Christian Hammers) 1202 postgresql-7.4 468 308 37 123 0 (Martin Pitt) 1531 libmysqlclient10 3277 214 518 48 2497 (Steve Langasek) 2253 postgresql-client-8.0 185 110 16 59 0 (Martin Pitt) 2261 postgresql-8.0 172 109 17 46 0 (Martin Pitt) 2332 mysql-client-5.0 120 102 1 17 0 (Christian Hammers) 2757 mysql-server-5.0 113 77 3 33 0 (Christian Hammers) -- Martijn van Oosterhout <kleptog@svana.org> http://svana.org/kleptog/ > Patent. n. Genius is 5% inspiration and 95% perspiration. A patent is a > tool for doing 5% of the work and then sitting around waiting for someone > else to do the other 95% so you can sue them.
Вложения
Richard Huxton wrote: > Andrew Sullivan wrote: > >> And consulting companies exist exactly to parrot whatever the current >> conventional wisdom is. Gartner only came across as ill-informed in >> that report (or any other I've ever read by them) when I actually >> knew something about the technolgies in question. > > > Gartner seem to have the same problems as the wider press do with any > technical field - they don't know what they're talking about. When it > is reporting on who said what when, both are fine. When it's > discussing figures, they seem broadly OK. Once you need to get > technical they generally flounder. This is my point. Gartner doesn't have a great reputation involving product comparisons and areas like TCO, technological advantages, etc. They come across much more like semi-techie journalists than quality consultants. When you compare their studies with those of, say, the IDC, their methods seem opaque, and their conclusions difficult to verify or even outright wrong. Similarly CapGeminii comes across as "the Windows experts" so why would one take their word on something like Linux? The problem I am mentioning is that consulting firms exist to provide competent advice. When they are proven wrong, it becomes a black eye. Indeed, they could have made a better case against PostgreSQL by simply saying "this is a less-known solution that we have not had prior opportunity to study. This may present risks in terms of finding qualified administrators and support staff." And if one is going to parrot conventional wisdom, it makes sense to say so and say why. Even something like "Oracle is the default solution in this area and PostgreSQL seems relatively untested. We see that as a serious business risk simply because it seems to be an unknown quantity" would be more honest than what came out. This is a general sickness which exists in many of the larger consulting/analysis firms. I assume that this is different than the technical consulting firms such as Accenture, but I have little experience in this area. Though to be fair this is largely a problem with hiring firms to do things that they are not able to do. For example, if I want a market forecast or an analysis of the current market, it might make sense to hire the IDC, but hiring them to help me make a decision between say, Linux and FreeBSD is likely to be very uninformative. Hiring Gartner to do this comparison seems to me sort of like using NT4 to manage user accounts in a 20000 user business (this is hardly uninformed: I am an MCSE). Best Wishes, Chris Travers Metatron Technology Consulting
Tom Lane wrote: > Christopher Kings-Lynne <chriskl@familyhealth.com.au> writes: > >>Strangely a pgsql to oracle exporter is a good thing. It'd be a great >>feature of PostgreSQL. Imagine how many people would start on >>PostgreSQL if they KNEW that one day they could easily move to Oracle if >>they needed to. Risk management. > > > Problem is: to offer such a thing with a straight face, we'd have to > confine ourselves to an Oracle-subset version of SQL. For instance, > lose the ability to distinguish empty-string from NULL. Oh please PLEASE *PLEASE* don't bend that way. Oracle has some SQL non compliant flaws at least one is serious: The inability to distinguish between the absence of value and an explicitly empty string is just ONE of Oracle's ridiculous fubarness. People who know what a NULL really is and use it properly have to program around Oracle's stupidity to "dumb it down" for the weak application developer, let's not do that. Terry > > regards, tom lane > > ---------------------------(end of broadcast)--------------------------- > TIP 5: don't forget to increase your free space map settings > -- Terry Fielder terry@greatgulfhomes.com Associate Director Software Development and Deployment Great Gulf Homes / Ashton Woods Homes Fax: (416) 441-9085