Обсуждение: Re: Is PostgreSQL ready for ...
I think the problem is that nobody wants to say "Oh yeah, solid as a rock!" and then have someone die as a result of a database error. However, I have personally witnessed unrepeatable database errors in both MSSQL and Oracle. Those databases both have marketing departments widely touting them as "mission critical" databases. I've had MSSQL databases fill the log tables with crap to the point where the whole database crashes because Microsoft decided it was time for a dll change -- and they didn't tell anyone "Hey, be sure to un-check the 'Let Microsoft ruin your life' checkbox". (that was an emergency personnel tracking database on an offshore oil rig). I've managed the development of an entire application based on Y2K compliant Oracle only to find out that it's only Y2K compliant if you enclose EVERY SINGLE F***ING READ AND WRITE in a goofy-ass "FORMAT" statement. Although, if you don't use the FORMAT statement, it will still accept four digit year entries without barking, and it will also display four digit years based on the two digit year it actually stores (so I guess if you never know the difference, what does it matter, right?). By the way, even if you use the Format statement, Oracle still craps out after 2035. (that application was a safety statistics tracking database). Soooo, for my part, I now put my faith in a database whose proponents do not face financial ruin if they tell you the truth. I much prefer a "Not yet implemented" message from the database, or a "this is still a little wonky" message from the developer, I much prefer that rather than the confident assurances you receive from Microsoft and Oracle. Especially when you pass on those confident assurances to your client and wind up looking like an ass. Sorry to explode. I feel better... David Boerwinkle
Bah, no comment on Microsofts reliability :) But Oracle I have used extensively 7.x versions. I implicitly trust its reliability...except I heard of some probs with version 8 when it came out :) Cant imagine someone using the same version of the database for 35 years tho. By that time you would have upgraded I would imagine. And Oracle upgrades its data types with no probs... ----- Original Message ----- From: <davidb@vectormath.com> To: <pgsql-general@postgreSQL.org> Sent: Monday, November 22, 1999 9:51 PM Subject: [GENERAL] Re: Is PostgreSQL ready for ... > emergency personnel tracking database on an offshore oil rig). I've managed > the development of an entire application based on Y2K compliant Oracle only > to find out that it's only Y2K compliant if you enclose EVERY SINGLE F***ING > READ > AND WRITE in a goofy-ass "FORMAT" statement. Although, if you don't use the > FORMAT statement, it will still accept four digit year entries without > barking, and it will also display four digit years based on the two digit > year it actually stores (so I guess if you never know the difference, what > does it matter, right?). By the way, even if you use the Format statement, > Oracle still craps out after 2035. (that application was a safety
At 08:51 PM 22-11-1999 -0600, davidb@vectormath.com wrote: >tracking database). Soooo, for my part, I now put my faith in a database >whose proponents do not face financial ruin if they tell you the truth. I >much >prefer a "Not yet implemented" message from the database, or a "this is >still a little wonky" message from the developer, I much prefer that rather >than the confident assurances you receive from Microsoft and Oracle. >Especially when you pass on those confident assurances to your client and >wind up looking like an ass. Yeah! I want truths. I'm a technical guy, I didn't get good technical knowledge by believing lies and fuzzy stuff. Thing is, Postgres could go various ways depending on the goals. If the priorities include stability and reliability, that's what you get. If the priorities are features at any cost, you get junk. Though Open Source projects are less susceptible to featuritis, they're far from immune. Trouble is many PHBs only compare stuff feature by brochure feature. Hmm, I must remember to put in "Maximum of one restart/reboot per year subject to clause X" in tender specs. Cheerio, Link.
> If the priorities include stability and reliability, that's what you get. > If the priorities are features at any cost, you get junk. > > Though Open Source projects are less susceptible to featuritis, they're far > from immune. Trouble is many PHBs only compare stuff feature by brochure > feature. We only do 2-3 major releases a year for a reason. If it is not reliable, it is useless. This is not a computer game. -- Bruce Momjian | http://www.op.net/~candle maillist@candle.pha.pa.us | (610) 853-3000 + If your life is a hard drive, | 830 Blythe Avenue + Christ can be your backup. | Drexel Hill, Pennsylvania 19026
At 01:18 AM 23-11-1999 -0500, Bruce Momjian wrote: >> If the priorities include stability and reliability, that's what you get. >> If the priorities are features at any cost, you get junk. >> >> Though Open Source projects are less susceptible to featuritis, they're far >> from immune. Trouble is many PHBs only compare stuff feature by brochure >> feature. > >We only do 2-3 major releases a year for a reason. If it is not >reliable, it is useless. This is not a computer game. Yep. Glad to hear that. My boss asked "MSSQL or Postgres" and a colleague and I said "Postgres". I figured we'd have fewer problems with Postgres, so what if we couldn't point fingers at someone else, better to get things done/fixed. Still good to hear that reliability is high on your list. Another colleague, was an Oracle guy and was nervous about Postgres - coz if anything goes wrong he may have to help :). But Oracle was way too pricey- we could have bought a house at the price they gave us... In contrast: Netscape proudly proclaimed that when it was a choice between features and stability, features always won. Explains a lot. Cheerio, Link.
> At 01:18 AM 23-11-1999 -0500, Bruce Momjian wrote: > >> If the priorities include stability and reliability, that's what you get. > >> If the priorities are features at any cost, you get junk. > >> > >> Though Open Source projects are less susceptible to featuritis, they're far > >> from immune. Trouble is many PHBs only compare stuff feature by brochure > >> feature. > > > >We only do 2-3 major releases a year for a reason. If it is not > >reliable, it is useless. This is not a computer game. > > Yep. Glad to hear that. Read our development history on our web site. It discusses this. -- Bruce Momjian | http://www.op.net/~candle maillist@candle.pha.pa.us | (610) 853-3000 + If your life is a hard drive, | 830 Blythe Avenue + Christ can be your backup. | Drexel Hill, Pennsylvania 19026