Ad­den­dum: By "fron­teers" I don't mean the dutch pro­fes­sion­al or­ga­ni­za­tion "Fron­teers", but more gener­ic peo­ple who very pas­sion­ate­ly try to make the im­pos­si­ble pos­si­ble by push­ing the en­ve­lope of web stan­dards in a pi­o­neer­ing, front run­ning, fron­tenders way of us­ing HTML, CSS and JavaScript in their strug­gle with the browsers. All I re­mem­ber is ppk in­vent­ing this la­bel that cov­ers this area ex­act­ly how I see it, but ap­par­ent­ly around Am­s­ter­dam the term has since gained more pop­u­lar­i­ty as the name of that or­ga­ni­za­tion. (Ad­den­dum 2: If I want to re­fer to that or­ga­ni­za­tion I'll use the term "bunch of un­pro­fes­sion­al HTML mon­keys" in­stead.)

The prob­lem with some front end de­vel­op­ers is that they be­lieve every­thing that is re­lat­ed to HTML, CSS and JavaScript should be aimed at mak­ing their job eas­i­er, be­cause it's the hard­est job in the world. They see stan­dards as sole­ly writ­ten for their line of work, to­tal­ly ig­nor­ing the fact that there are ac­tu­al­ly real de­vel­op­ers who have to im­ple­ment those stan­dards into a brows­er first, and not every sit­u­a­tion the stan­dards have to cov­er has the con­ve­nience of up­grad­ing to the lat­est mozil­la re­lease can­di­date to fix some is­sues.

Con­sid­er the fol­low­ing sce­nario. Some man­u­fac­tur­er of a piece of hard­ware wants to have a fan­cy user in­ter­face on the thing. The piece of hard­ware could be a tele­vi­sion set, a fridge, in­dus­tri­al clean­ing equip­ment, what­ev­er. The man­u­fac­tur­er can slap a lit­tle LCD touch screen on it and put a cheap board in it that pro­vides the user in­ter­face. In a per­fect world, shouldn't that user in­ter­face be build with open stan­dards like HTML, CSS and JavaScript? So the man­u­fac­tur­er sets out to em­bed a brows­er in the thing. The man­u­fac­tur­er can buy the re­quired soft­ware com­po­nents that are al­ready avail­able and put them in. But be­cause the em­bed­ded brows­er is lim­it­ed to that spe­cif­ic de­vice it could take ad­van­tage of fea­tures spe­cif­ic to that de­vice. The em­bed­ded brows­er can be ex­tend­ed to han­dle CSS that isn't stan­dard so the de­vel­op­ers of the user in­ter­face have a few ex­tra prop­er­ties to eas­i­ly im­ple­ment the spe­cif­ic fea­tures.

The CSS stan­dard takes care of this by al­low­ing ven­dor pre­fix­es. Without ven­dor pre­fix­es there is a chance that a new­er ver­sion of the stan­dard would in­tro­duce a prop­er­ty that has the same name as some­thing the ven­dor has in­vent­ed, caus­ing a con­flict and con­fu­sion. So ven­dor pre­fix­es are a good thing.

We can't get away with gener­ic -draft- or -beta- pre­fix­es be­cause our man­u­fac­tur­er could then im­ple­ment their fea­tures as a -beta- prop­er­ty, and the up­stream provider of the CSS en­gine for the em­bed­ded brows­er could sud­den­ly also in­tro­duce that same -beta- prop­er­ty for some­thing else, caus­ing the same con­flict­ing sit­u­a­tion that could oc­cur with­out ven­dor pre­fix­es. So the up­stream provider should use some­thing like -up­stream­provider- and the man­u­fac­tur­er some­thing like -man­u­fac­tur­er- as ven­dor pre­fix. And when the man­u­fac­tur­er then sells the code of it's em­bed­ded brows­er to an­oth­er man­u­fac­tur­er they can do the same thing and every up­date from the up­stream provider pass­es the work­flow with­out prob­lems.

The only prob­lem is that cur­rent­ly a large part of the browsers used for surf­ing the web on desk­top and mo­bile de­vices is based on open source en­gines, which makes it pos­si­ble for web de­vel­op­ers to find out what all their ven­dor spe­cif­ic fea­tures are and what they do. Adopt­ing all these ven­dor spe­cif­ic fea­tures to do cool things be­fore they be­come a prop­er stan­dard - and the com­pe­ti­tion does the same cool things - web de­vel­op­ers walk straight into the trap that was fu­el­ing the brows­er wars in the 90s. Ven­dor pre­fix­es aren't de­fined as be­ing la­tent stan­dards and it's just wish­ful think­ing to see drop­ping them as a so­lu­tion for eas­i­er web de­vel­op­ment, or speed­ing up some stan­dariza­tion process. But ap­par­ent­ly web de­vel­op­ers have to waste the time they gained from drop­ping sup­port for In­ter­net Ex­plor­er 6 by set­ting up the next self in­duced mess that makes their job so hard.