Mwanji Ezana inquired me in a remark on my previous blog post
I consider one of the major vantages of ORMs is their lazy-loading, caching and query-batching ability. It Holds not simply about generating a scheme and enquiries.
Altogether your anti-ORM stations, I 've ne'er seen you name these capablenesses. What make you do of them? Are they unimportant to you?
( Thanks for reading, Mwanji! )
I hold a couple of inputs, and I 'll need to get with the definition of ORM itself: as I discourse in my previous blog post , ORM holds come to intend a great deal of things, including SQL plotters like iBATIS, which I personally make n't view ORM, but which, apparently, much/some of the programing community makes.
So I 'll iterate that I merely dislike the ORMs that compose SQL for me behind my dorsum. For the remainder of this blog entry, if I tell ORM, only believe ORM of the form that automagically makes things for you; not ORM that alleviates easier authorship of SQL inquiries, such as iBATIS.
When it comes to lazy-loading, caching, and query-batching, I wish them all! It Holds merely that I consider they are all best when uncouple from ORM. Caching, particularly, is arguably something that you may desire uncouple from your ORM.
E.g., Django, holds separate ORM and caching mechanisms: caching even holds its ain standalone chapter
in the Django book.
I conceive this add up: caching is for more merely database questions, so I conceive it Holds a great installation to offer outside of an ORM.
I view query-batching to be another installation that can even as easily be offered outside of an ORM. If anything, the most effectual shipways I cognise to make big batch businesses on RDBMSs is to apply the tools offered by the RDBMSs to make so, instead than those offered by any ORM or library. RDBMSs ' batch tools usually work best from the bid line, permiting you to not simply avoid your ORM for batch occupations, but avoid the whole application whole.
I believe I 'm partially anti-ORM for esthetic grounds: I essay to avoid resistance mismatches alternatively of embracing them.
View a task where an object framework deadly draws the concern arena, but there Holds a demand to store the data in SQL, mayhap because of SQL 's ability to generate great studies, or whatever. It Holds the classical object/relational resistivity mismatch.
Some ( most? ) tasks comprehend this mismatch by utilise some rather ORM to bridge the spread.
Personally, I 'd see if I could happen a really good object database so that I maked n't need to plow with the mismatch. Why not store my objects in an object database and not even need to handle with an RDBMS? Mayhap there Holds an object database out there that can still generate the studies I ask, or make other things I though I necessitated SQL to make.
But then, if it turned out that the task required features merely SQL can supply, I 'd consider long and hard about whether or not my concern data rattlingly
need to be an object framework. Perhaps I could apply the relational information framework after all? Peradventure in my application codification I could employ lists of maps to correspond my data ( so it would be very much like SQL tables and/or outcome sets ) and avoid the resistance mismatch by essentially taking my relational framework up into my application bed.
But that Holds only Maine I 'm heavily biased towards licking jobs by not shoulding lick them in the first spot: Got an resistivity mismatch? Pick a side. Now you make n't need to intercede between two slipways of looking at the same data, because you but extinguished one by determining that the other was more important.
But I 'm happy to acknowledge to at least two things:
1 ) Not all developers hold the elimination/simplicity preconception that I hold, and
2 ) Not all undertakings can only pick one paradigm and extinguish another.
The continued popularity of ORM must intend that a great deal of people are utilizing it in a good deal of successful labor.
I may personally surmise that a good deal of tasks belike win in malice of
ORM, but I need to acknowledge that it Holds justly because ORM holds ne'er been a good scene for the undertaking I 've worked along, so that experience holds worked my thought. But I conceive us coder involve to acknowledge that it Holds a large existence of programing jobs out there, and one size makes not accommodate all.
If your task is making great, and you 're utilizing ORM, so in the context of your labor, you are right, and I am incorrect. I 'm glad you 're not listening to my criticisms of ORM, and I 'm glad you 're stick to what plants.
But then, if ORM is not working for you, or it Holds presenting some strain, check out Ted Neward 's The Vietnam of Computer Science
, or about 25 % of the blogs I 've ever pent. ; - ) You may happen some observations that peal true, even if you proceed to apply ORM.
Related posts:
Season of Mists and Blaze Sunlight
Figure 5: Vertical Trip Tych
Soap Roach 's M'Boom
![[info]](http://l-stat.livejournal.com/img/userinfo.gif){kind=small}
